Showing Posts From

Technology

Seeing opportunities with AI

Artificial Intelligence (AI) is changing the way businesses operate, offering new opportunities and challenges. As a C-level executive, it's important to understand how AI can benefit your company while managing the risks involved. Setting Your AI Goals First, you need to decide what you want to achieve with AI. Do you want to use it to improve internal processes or to create new products and services? Your ambition will guide your strategy and set realistic goals. For example, AI can help streamline back-office tasks, making them faster and more efficient. Or, you might use AI to offer personalized customer experiences, which can lead to higher customer satisfaction and loyalty. Choosing the Right Approach Next, consider how you will implement AI. There are different ways to do this. You can use pre-built AI tools that are already available. This is quick and doesn’t require much technical knowledge, but it may not fit your specific needs perfectly. Alternatively, you can adapt existing models with your own data to make them more tailored to your business. This approach is more flexible but requires more expertise. Lastly, you can develop your own AI system from scratch. This gives you full control but is more expensive and time-consuming. Choosing the right path is crucial. It affects how quickly you can start using AI and how much it will cost. For instance, if your goal is to quickly improve customer service, a pre-built solution might be the best choice. If you need a highly customized solution for a specific problem, developing your own AI system might be necessary. Navigating the Risks Using AI also comes with risks. These include unreliable outputs, data privacy issues, cyber threats, and regulatory concerns. For example, AI systems can sometimes produce incorrect or unexpected results. This can happen if the data used to train the AI is flawed or if the system encounters new situations it hasn’t seen before. Ensuring data privacy is crucial, especially when handling sensitive information. You need to comply with regulations like GDPR in Europe or HIPAA in the U.S. Cyber threats are also a concern. AI systems can be targeted by hackers, putting your data at risk. This means you need to have robust cybersecurity measures in place. Additionally, different countries have different rules about AI, and you need to follow them. This can be complex, as regulations can change quickly and vary widely. For instance, regular audits and compliance checks can help ensure you stay within legal boundaries. Leading with Vision and Prudence Leading with AI requires a balanced approach. You need to support innovation while also ensuring safety and ethical considerations. This involves engaging stakeholders, balancing speed and caution, and fostering a culture of learning. Engaging stakeholders means talking to everyone involved, from developers to end-users, to get their input and support. This helps build a sense of ownership and alignment. Balancing speed and caution is also important. You need to move fast to stay ahead of competitors but take time to ensure your AI is reliable and secure. Fostering a culture of learning means encouraging your team to learn about AI and keep up with new developments. This helps keep your organization ahead of the curve. Wrapping Up AI offers a unique chance for C-level executives to drive growth and innovation. However, it also presents significant challenges. By carefully planning and managing risks, you can use AI to improve your business and stay ahead of the competition. In summary, leading with AI means setting clear goals, choosing the right deployment strategy, and being prepared for risks. With the right approach, you can unlock the full potential of AI for your organization. True leadership means guiding your company through the complexities of AI with vision and resilience.

Co-managed IT explained: who is really responsible?

Choosing how to run your IT infrastructure is one of the most important strategic decisions a business can make. However, many business leaders struggle with confusing terminology in the IT service provider landscape. Terms like co-managed IT, co-sourcing, fully managed services, and co-creation are often used incorrectly, leading to failed partnerships and unclear expectations. Understanding what these models actually mean, how responsibilities are divided, and how financial billing works is essential before signing any contract. The landscape of IT management models To make informed choices, business leaders must clearly distinguish between the different ways IT services can be delivered and organized. Under an insourcing model, a business handles all technology needs internally by hiring and managing its own personnel. Outsourcing, by contrast, transfers an entire process or department to an external provider who guarantees specific performance targets. Co-sourcing takes a staff augmentation approach by bringing in external personnel to work under your internal team's direction, adding temporary capacity without shifting operational control. Service delivery models also differ in scope and management approach. A standard managed service focuses on buying a specific functional outcome under a strict agreement, while remote managed services rely on software tools to monitor systems from a distance. Fully managed services go a step further by handing over complete operational responsibility for the entire IT environment to an external partner. Finally, co-managed IT involves an internal team and a provider managing a domain together, whereas co-creation focuses on jointly developing new digital products rather than managing existing systems. Deep dive into co-managed IT: what it is and what it is not Co-managed IT is often misunderstood in the service provider market, where it is frequently confused with buying extra staff or single software tools. In reality, a true co-managed setup is a joint operational partnership. Both the internal IT team and the external provider actively manage a specific domain together by sharing access to management platforms, support queues, and daily workflows. Both parties share equal accountability for system health, overall uptime, and cybersecurity. This approach is fundamentally different from other sourcing arrangements. It is not co-sourcing because co-sourcing merely supplies extra hands without transferring operational accountability to the vendor. It is also distinct from co-creation, which develops new intellectual property, and traditional outsourcing, which removes the internal team from daily operations entirely. Companies select co-managed models when they have a capable internal team that understands the business, but needs enterprise-grade tools, 24/7 coverage, and specialized knowledge. Financially, co-managed services usually rely on a predictable monthly fee per user or device, combined with set rates for project support. Deep dive into co-creation: what it is and what it is not Co-creation is another term that is often misused when organizations confuse custom software development with operational IT management. At its core, co-creation is a collaborative development strategy where a client and a technology vendor build a software tool together. The client provides domain expertise, practical feedback, and operational requirements, while the vendor contributes technical architecture, software engineering, and scalable infrastructure. This model should not be confused with standard custom software development, where a client pays the full cost to keep exclusive rights. Nor should it be mistaken for co-managed IT or co-sourcing, as co-creation focuses on building new digital tools rather than supporting daily IT operations. Businesses choose co-creation when standard commercial software falls short, but building custom tools alone is financially unfeasible. Financially, the client typically receives lower development rates or early software access. In return, the vendor retains the core intellectual property and creative freedom, allowing them to market and sell the solution to other commercial customers. The shared responsibility model: operational versus legal reality When working with an external IT partner, dividing responsibilities correctly is critical to avoiding operational gaps and legal surprises.IT Sourcing Model Operational Execution Operational Responsibility Legal Accountability Common Billing StructureInsourcing Internal staff Internal IT management Internal business board Internal salaries and capital spendOutsourcing External provider External service provider Internal business board Fixed monthly contract or service feeCo-sourcing Internal staff & external personnel Internal IT management Internal business board Time and materials or daily ratesCo-managed Shared internal and external team Joint shared responsibility Internal business board Fixed fee per user/device + project rateCo-creation Joint development team Joint development leadership Internal business board Discounted dev fees + IP retentionFully Managed External provider External service provider Internal business board Fixed monthly fee per user or deviceOperationally, you can delegate tasks and share daily responsibilities with a partner. In a co-managed environment, the vendor might handle backup management and software patches while your internal team supports end users. If a backup fails due to vendor negligence, the vendor is operationally accountable based on agreed service levels. However, legal responsibility works very differently. Regulators and courts hold your board of directors legally accountable if a cyberattack occurs or privacy laws are violated. While you can seek financial damages from a partner for breach of contract, ultimate legal accountability remains with your business. Closing thoughts Modern IT management requires a clear understanding of where effort ends and true responsibility begins. Misidentifying your sourcing model leads to operational confusion, unfulfilled promises, and unmanaged business risk. By defining roles, financial structures, and legal boundaries early, organizations can build effective partnerships that protect their operations. True IT partnerships are built on shared operational accountability, but business leaders must remember that legal responsibility can never be outsourced.

Governing AI agents while driving business value

Technology leaders today face a very difficult choice. On one hand, company executives want to see clear financial results from artificial intelligence investments. On the other hand, using automated AI tools introduces safety and security risks that older systems never had. To handle this successfully, companies must move away from simple testing and focus on clear rules, cost control, and practical learning. Higher pressure on budgets and financial results For a few years, many companies spent money on AI just to see what it could do. Today, that period of open spending is over because business leaders want to see real value. While many companies are still spending more money on technology, only a small number expect to get fast returns on their AI investments. This gap between spending and real results puts a lot of pressure on technology directors. To solve this, successful companies are changing their approach in three main ways:Focusing on clear tasks: Leaders are stopping general pilot projects that have no clear goals and are choosing tasks where results are easy to measure. Setting goals before starting: Good teams decide on clear targets before launching a project, so they can prove the financial benefits later. Managing hidden costs: Using AI models too much, paying high usage fees, and running uncontrolled software tools can quickly become too expensive.The hidden risks of automated software tools The step from standard AI models to automated AI agents creates new risks for companies. Standard tools just answer questions, but automated agents can run code, change databases, and complete complex actions across different systems by themselves. If these systems operate without strict rules, they can make unexpected mistakes, like accidentally deleting important company databases. At the same time, many employees are using unapproved AI tools on their own. Workers in different departments often use personal accounts or free online tools to do their jobs faster. While this can save time, it can also leak private company information and create serious security problems. Changing safety rules from yearly checks to daily monitoring Old ways of managing software risks, such as checking rules once a year, do not work for fast AI systems. Because automated tools work continuously and very quickly, security plans must adapt to monitor them all the time. To keep systems safe without stopping work, technology managers should follow a clear plan:Limiting system access: Automated tools should never have full access to everything; their permissions must match the exact task they are doing. Creating strong central rules: Instead of changing safety settings for every new tool, create one strong system that decides what data can be used and when a human must check the work. Keeping complete activity logs: Every action taken by an automated tool must be saved in a list so managers know what happened and why.** Giving clear responsibility to staff: Set up mixed teams and clear ownership so that technology, legal rules, and business goals work together.Helping employees learn and adapt As software work becomes more automated, companies face a new human problem. Younger workers and junior developers who rely too much on AI tools might not learn basic building skills. If they do not learn from real mistakes, it becomes hard for them to notice when an AI system gives a wrong answer. To fix this problem, business leaders need to build a learning culture. Experienced staff members should guide younger workers, encourage open discussions about technical issues, and check AI outputs carefully. Good training across the whole business helps everyone understand both the power and the limits of these new tools Closing thoughts Navigating the complex world of modern technology requires both fast innovation and careful control. Successful leaders will not be the ones who buy every new tool, but those who build clear safety rules, manage spending carefully, and stay responsible for their automated systems. True progress in technology happens when we combine speed with total responsibility.

Large vs. Small Language Models: Understanding the differences and choosing the right tool

In the early days of generative artificial intelligence, tech companies believed one main rule: bigger is always better. Massive AI models like OpenAI's GPT-4, Google's Gemini Ultra, and Meta's Llama-3 proved that adding hundreds of billions of parameters unlocked incredible skills. These models could solve complex logic problems, write software, and translate languages easily. However, a new trend is taking over the tech world: Small Language Models (SLMs). Models such as Microsoft's Phi-3, Google's Gemma, and Meta's Llama-3-8B show that smaller models can also be smart, fast, and much cheaper to use. To understand the AI landscape today, models are generally divided into three main categories:Large Language Models (LLMs) [70B+ Parameters]: Massive models trained on huge amounts of internet data. They require powerful cloud servers to run and act as general-purpose experts. Medium Language Models [13B to 70B Parameters]: Balanced models that offer strong reasoning skills while still being easier for companies to host privately. Small Language Models (SLMs) [1B to 10B Parameters]: Compact models designed to run efficiently on small hardware, such as regular laptops, smartphones, or small internal company servers.Hardware constraints: Why model size matters To understand why SLMs are becoming so popular, we need to look at computer hardware, specifically graphics memory (VRAM) and speed. Memory requirements (VRAM) To run an AI model, its weights (parameters) must be loaded directly into a computer's high-speed graphics memory.A large 70-billion parameter model needs around 140 GB of VRAM to run at standard quality. This requires high-end enterprise hardware costing tens of thousands of dollars. In contrast, a small 8-billion parameter model can be compressed (quantized) to run using less than 5 GB of VRAM. This means it can easily run on a standard work laptop or a modern smartphone.Speed and latency Big models need to move massive amounts of data back and forth through hardware every time they generate a word. Smaller models carry much less data, which allows them to generate text much faster. This makes SLMs ideal for real-time tasks like live customer chat or typing assistance. How small models get so smart How can a small model perform almost as well as a giant model from a few years ago? The secret lies in high-quality data and smart training techniques. Traditional LLMs learn from raw internet text (billions of webpages, social media posts, and slang). Modern SLMs, on the other hand, learn from curated, high-quality "textbook" data and simplified lessons from larger models.Filtered Synthetic Data: Instead of learning from random internet chatter, modern SLMs are trained on clean, high-quality data created by larger AI models. This includes clear coding examples, textbooks, and step-by-step logic exercises. Knowledge Distillation: This is a process where a large "Teacher" model helps train a smaller "Student" model. The student learns to copy the reasoning patterns of the teacher without needing the giant memory size.Key differences at a glanceFeature Large Language Model (LLM) Small Language Model (SLM)Model Size 70B to 1 Trillion+ parameters 1B to 10B parametersHardware Needed Massive enterprise GPU servers Standard laptops, phones, single GPUsMemory Footprint Very High (100GB+ VRAM) Low (2GB to 10GB VRAM)Response Speed Slower for large answers Extremely fast generationOperating Cost High cloud API or server fees Very cheap to host locallyData Privacy Data usually sent to the cloud Data can stay fully on your local deviceCombining SLMs with company data (RAG) Many organizations assume they need a giant AI model to understand their company's internal files. However, using an AI model as a giant memory bank is inefficient and often leads to false answers (hallucinations). Instead, smart companies combine Small Language Models with a system called Retrieval-Augmented Generation (RAG):Step 1: User asks a question. Step 2: The system searches internal company documents for the facts. Step 3: The exact document text is given to the SLM. Step 4: The SLM reads the text and writes a clear answer.Because the SLM does not need to memorize all company facts inside its parameters, a small 8B model paired with RAG often outperforms a large, expensive LLM at a fraction of the cost. When to choose an LLM vs. an SLM Choosing the right model depends on your specific goals, budget, and privacy requirements. Choose a Large Language Model (LLM) when:You need complex reasoning: Writing complicated software code, analyzing vague legal documents, or solving advanced scientific problems. You build autonomous agents: AI systems that need to plan multiple steps and interact with external tools independently. Your queries are unpredictable: Your application covers many completely different subjects without a fixed focus.Choose a Small Language Model (SLM) when:Speed is critical: Applications like real-time translation, autocomplete, or instant customer support. Privacy is mandatory: Healthcare, finance, or legal tasks where data cannot leave the local building or device. You operate on a budget: Running high volumes of daily requests without paying expensive cloud API subscription fees.Closing thoughts Artificial intelligence is no longer just about building the largest possible model. While giant LLMs remain important for cutting-edge research and complex logic, Small Language Models are proving to be the most practical choice for daily business operations. By using clean training data, clever optimization, and targeted document systems, SLMs deliver fast, private, and cost-effective performance. The best AI architecture is not about using the biggest model available, but finding the smallest model that can solve your problem effectively.

The AI productivity paradox: Why more tools aren't saving us time

Artificial intelligence has spread faster than almost any other technology in human history. Today, workers across every industry use generative AI daily. They use it for writing reports, designing presentations, writing software, and summarizing long meeting notes. Major software companies have embedded AI directly into our email clients, office suites, and project dashboards. On paper, this should save us hours of work every week. Yet, if you look at modern business statistics, overall productivity has barely moved. Many business leaders are left asking the same frustrating question: If everyone is using AI, why is work not getting done any faster? This situation is not actually new. It is a modern version of the famous "productivity paradox" observed by economist Robert Solow in the 1980s. Back then, he noted that computers were visible everywhere except in the economic productivity numbers. Today, AI faces the exact same challenge. Why AI saves minutes, not whole processes The main reason for this productivity gap is simple: most people use AI to speed up small, isolated tasks rather than fixing full workflows. For example, a customer service agent might use AI to draft a quick reply to an email. The drafting takes five seconds instead of five minutes. However, that message still needs manual review, manager approvals, and input into old database systems. The bottleneck simply moves to another part of the process. In addition, several hidden time-wasters prevent AI from delivering its full potential:The Double-Checking Burden: AI outputs are rarely perfect on the first try. Employees end up spending significant time checking facts, correcting hallucinations, and editing formatting. Tool Overload: Organizations often use multiple specialized AI tools at the same time, such as ChatGPT, Claude, Midjourney, and GitHub Copilot. Deciding which tool to use and switching between them creates mental fatigue. The "More Content" Trap: Because creating documents and emails has become easier, companies generate much more of them. This creates a massive ocean of reports and emails that other employees must spend time reading. Constant Context Switching: Workers constantly jump between Slack messages, email, AI chats, and project boards, which drains mental energy throughout the day.What history teaches us about real efficiency MIT economist Erik Brynjolfsson points out that groundbreaking technologies rarely boost productivity immediately. He compares the current adoption of AI to the arrival of electricity in factories during the late 19th century. When factory owners first replaced steam engines with electric motors, productivity did not go up right away. It was only when they completely redesigned factory layouts and assembly lines around electricity that output exploded. Old Approach: [Standard Process] + [Add AI Tool] = Minimal Time Saved New Approach: [Redesigned Process Built for AI] = Massive EfficiencySimilarly, Wharton professor Ethan Mollick emphasizes that AI works best as a collaborative partner rather than a basic tool. Companies that see massive productivity gains do not just give their workers an AI login; they fundamentally rethink how work gets done. Real-world example: Support & software developmentCustomer Support: Instead of using AI just to suggest email templates, leading companies let AI agents sort tickets, handle routine queries autonomously, and route complex edge cases directly to human experts. Software Engineering: Rather than using AI merely to write single lines of code, teams integrate AI across the whole cycle—from initial architecture planning and automated unit testing to security checks and documentation.Looking ahead: The shift to autonomous agents We are currently moving from simple AI assistants toward autonomous AI agents. New multimodal agentic systems—like Alibaba’s Qwen 3.7 Plus—can look at user interfaces, click buttons, navigate websites, and complete multi-step tasks across different software without constant human prompting. As these tools mature and become affordable to deploy, the central question for businesses will change. It will no longer be "Should we use AI?" but rather "How must we redesign our work to let AI perform whole tasks effectively?" Closing Thoughts Having access to the most powerful AI tools in the world will not automatically make your team faster or smarter. Technology only provides the raw capability; real success depends on how thoughtfully you restructure your daily habits, workflows, and organizational structures to support it. True productivity in the AI era is not about doing old tasks faster. It is about designing completely new ways of working.

Stop using engineers as shock absorbers

Look at almost any modern tech job advertisement today, and you’ll see the exact same list of benefits: “Flexible hybrid work, autonomous culture, latest hardware, and regular team events.” Yet, despite these perks, tech companies worldwide face a persistent crisis: their highest-performing senior engineers and tech leads are silently walking out the door. Industry data confirms this gap. According to global developer experience benchmarks, over 65% of senior engineering turnover is driven by organizational friction and administrative noise, rather than technical difficulty or compensation. Developers do not quit because of a lack of team socials or fruit baskets. They leave when their day-to-day job becomes buffering their team against executive indecision, sitting in low-value alignment meetings, and blunting structural chaos. The "Human shock absorber" In many growing software organizations, a subtle leadership failure occurs as teams scale. When executive boards struggle to establish clear strategic boundaries or resolve cross-departmental friction, they quietly delegate that responsibility downward. They create what can only be called a Frankenstein Role: a Tech Lead or Staff Engineer who is asked to be 100% hands-on architect, 100% people coach, and 100% process firefighter. Instead of solving complex technical problems or building scalable cloud architectures, your highest-paid technical experts become human shock absorbers. They spend up to half their working week absorbing leadership noise, translating vague goals, and mediating conflicts that should have been settled at the C-level. Research on developer cognitive load shows that modern software engineers spend less than 30% of their actual workday writing code or designing software. The remaining time is consumed by context-switching, status updates, and navigating organizational friction. The real math behind senior engineering turnover When a burned-out senior engineer or lead resigns from a bloated role, the financial damage on the P&L statement is far larger than most executives realize. The true total cost of losing a key technical figure can be broken down using standard engineering talent benchmarks:Cost Category Impact Level DescriptionDirect Replacement Costs Significant Agency fees, interviewing hours, sign-on packages, and competitive market salaries for senior talent.Onboarding & Ramp-Up Substantial Lost productivity during the 6 to 9 months it takes a new senior engineer to master a complex codebase.Contagion Effect (Domino Turnover) High Risk McKinsey research shows that when a respected lead quits, team members are up to 35% more likely to leave within 6 months due to increased workload and chaos.Roadmap & Market Delay Severe Slid delivery dates, delayed feature releases, and missed market opportunities.Plaguing your organization with high turnover isn't a recruitment issue—it is a direct leadership leak. You can't code out of broken governance With the rapid adoption of AI coding assistants, agentic dev-tools, and automated testing suites across engineering teams, leadership teams often assume tech investments will solve their productivity bottlenecks. However, recent studies on AI engineering adoption highlight a clear contradiction: While AI assistant tools improve individual line-of-code generation by 15% to 20%, total organizational delivery velocity in chaotic companies improves by less than 3%.Why? Because generating code was never the primary bottleneck. If your decision-making process is slow, your boundaries are blurry, and your teams are misaligned, AI tools simply help your developers build the wrong things faster. Buying AI licenses to compensate for poor organizational design is one of the most expensive escape routes on an IT balance sheet. You cannot solve a structural leadership deficit with a software subscription. How to sanitize your engineering leadership Restoring execution speed and retaining top-tier engineering talent requires structural clarity at the top. You don't need another soft skills workshop, agile transformation, or internal culture initiative. You need clean leadership architecture: 1. Keep C-suite accountabilities at C-level Executives must set firm strategic priorities, establish binary boundaries, and clean up inter-departmental politics. Never ask a Tech Lead or Engineering Manager to resolve organizational friction without giving them explicit executive authority. 2. Ruthlessly separate technical roles from line management Stop expecting senior engineers to be elite software architects and full-time people managers simultaneously. Create clear, parallel career tracks:Individual Contributor (IC) Track: Focused 100% on architecture, technical execution, and code quality. Engineering Management Track: Focused on people development, resource allocation, and team enablement.3. Measure friction, not just output Instead of tracking raw output or ticket velocity, measure organizational friction:How many hours a week do senior leads spend in alignment meetings? How often do decisions made at the top get reopened three weeks later? How long does it take to get a clear 'yes' or 'no' on technical decisions?Closing thought Senior A-players in software engineering do not leave companies because the work is hard. They leave when the work is made unnecessarily chaotic by a lack of leadership structure. The next significant improvement to your bottom line and product delivery won't come from a new framework, a recruitment drive, or another AI tool. It will come from eliminating the hidden operational friction that is draining your lead engineers today. Ask yourself: Is your executive team providing clear boundaries for your engineers to build great products, or are you using them as human shock absorbers for organizational noise?

The ungoverned cloud: Why cloud strategies fail at execution.

In boardrooms across Europe, cloud strategy is undergoing a harsh reality check. For years, the narrative was centered on speed and migration. Today, executive teams face a very different set of challenges: unpredictable cloud expenditure, strict regulatory mandates (NIS2, BIO2, EU AI Act), and diffuse operational accountability. When external audits reveal that two-thirds of cloud environments lack proper control, the executive reflex is predictable: install a heavy Governance Board, write 80-page policy manuals, and require manual sign-offs for every change. This approach fails every time. It creates shadow IT, paralyzes delivery teams, and fails to eliminate actual risk. Personally, I view cloud governance not as a bureaucratic brake, but as an operational operating system. True governance provides clear guardrails, automated compliance, and organizational clarity—allowing engineering teams to move fast safely. To achieve this, organizations must move away from theoretical policies and implement a functional Cloud Center of Excellence (CCoE).The 5 pillars of cloud governance Before structuring your team, you must define what cloud governance actually encompasses. Mature cloud governance covers five distinct operational domain pillars:Pillar 1: Financial Management (FinOps)Shifting from static annual IT budgets to dynamic unit economics, continuous cost allocation, and real-time optimization.Pillar 2: Security & Regulatory Compliance (NIS2 / BIO2)Enforcing baseline controls aligned with NIS2, BIO2, ISO 27001, and GDPR across all cloud landing zones.Pillar 3: Automation & Platform EngineeringEliminating manual infrastructure configuration through Infrastructure as Code (IaC) and automated developer platforms.Pillar 4: Identity & Data Control (Zero Trust)Implementing Zero Trust architecture, strict least-privilege principles, and explicit data boundaries.Pillar 5: AI & Emerging Tech Governance (ISO/IEC 42001)Setting parameters for responsible AI use under the EU AI Act and ISO/IEC 42001, preventing unmanaged shadow-AI implementations.What is expected of C-level leadership? Cloud governance cannot be delegated away to IT or a compliance team. Real governance requires active C-level involvement, clear sponsorship, and strategic alignment. Here is what is explicitly expected of executive leadership across each core domain:C-Level Role Core Executive Expectation & Operational ResponsibilityChief Executive Officer (CEO) & Board Treat Cloud Governance as Risk Management: Recognize that cloud failure, data breaches, and non-compliance carry direct board liability under NIS2. Establish risk appetite boundaries and mandate cross-functional governance across the company.Chief Operating Officer (COO) Align the Operating Model & CCoE Mandate: Provide the CCoE with formal authority to set organizational standards. Break down functional silos between IT, Security, and Business units, ensuring that delivery speed never bypasses compliance.Chief Financial Officer (CFO) Enforce Financial Accountability (FinOps): Shift financial oversight from traditional CapEx IT depreciation to dynamic OpEx management. Demand unit-cost transparency and require Business/Product Owners to account for cloud consumption within their P&L.Chief Information / Technology Officer (CIO/CTO) Drive Modern Architecture & Enablement: Transition engineering teams away from manual ticketing towards self-service platforms (IDPs). Enforce "Policy as Code" and ensure cloud infrastructure aligns with architecture goals.Chief Information Security Officer (CISO) Automate Guardrails Over Gatekeeping: Move from reactive security reviews to proactive, automated policy enforcement. Integrate NIS2, ISO 27001, and AI compliance directly into deployment pipelines.Key Leadership Takeaway: Executive leadership is not expected to manage cloud settings or review code. Leadership is expected to set parameters, grant mandate, enforce accountability, and model the culture required for operational discipline.Enablement, not control The central execution engine of cloud governance is the Cloud Center of Excellence (CCoE). Too many companies misinterpret the CCoE as an architectural approval committee that meets every Thursday to review tickets. That is the quickest way to kill organizational momentum. Gatekeeper vs. EnablementAnti-Pattern: The Gatekeeper CCoE Modern Pattern: The Enablement CCoEManually reviews and approves architectural change requests. Builds automated guardrails and self-service templates.Writes static policy PDFs that engineers rarely read. Embeds policy directly into deployment pipelines (Policy as Code).Acts as a centralized bottleneck for cloud adoption. Functions as an internal product team serving delivery teams.Measures success by policy compliance and audit logs. Measures success by engineering velocity, security, and cost efficiency.Structure & core roles A successful CCoE is a lean, cross-functional team that brings together key domains. It does not replace engineering teams; it empowers them.Executive Sponsor (COO / VP Operations): Secures budget, aligns governance with corporate P&L goals, and resolves organizational friction between business units. Cloud Lead / Architect: Defines overall multi-cloud strategy, Landing Zone standards, and reference architectures. Cloud Security & Risk Specialist: Translates regulatory requirements (NIS2, ISO 27001, EU AI Act) into actionable security policies and automated checks. Platform Lead / Software Architect: Drives Platform Engineering, building Internal Developer Platforms (IDPs) and self-service "Golden Paths". FinOps Practitioner: Analyzes cloud consumption data, establishes unit-cost metrics, and works directly with product owners on cost accountability.Practical implementation: A 4-phase roadmap Implementing cloud governance across an organization requires a phased, practical approach. Phase 1: Establish the charter & landing zone architectureDefine the CCoE Charter: Formally declare the team's purpose, scope, and mandate across the business. Build Landing Zones: Create standard multi-account cloud structures (e.g., AWS Organizations or Azure Management Groups). Isolate workloads by environment (Dev, Test, Prod) and business unit. Implement Centralized Logging: Ensure audit trails, identity logs, and network traffic are automatically ingested into a central SIEM system from day one.Phase 2: Automate guardrails (Policy-as-Code)Define Preventive & Detective Controls: Use native cloud policies (e.g., Azure Policy, AWS Service Control Policies) to enforce mandatory constraints: Preventative: Block public S3 buckets or unencrypted storage volumes from ever being created. Detective: Automatically flag and alert security teams when a resource drifts from baseline configuration.Tagging Strategy Enforcement: Mandate metadata tags (Owner, CostCenter, Environment, DataClassification) at deployment time. If a resource lacks tags, auto-remediate or reject the build.Phase 3: Platform engineering & self-service (Golden Paths)Build the Internal Developer Platform (IDP): Provide engineering teams with a self-service portal (e.g., Backstage) to provision compliant infrastructure in minutes. Publish Golden Paths: Pre-package approved architectures (e.g., secure microservice deployment, compliant SQL cluster) that include security, monitoring, and backups by default. Community of Practice: Establish cloud guilds to train product teams, share best practices, and accelerate internal skills development.Phase 4: FinOps maturity & Responsible AI governanceShift-Left Cost Management: Integrate cost-estimation tools into CI/CD pipelines so developers see the estimated monthly bill before merging code. Establish AI Guardrails: Deploy private API endpoints for Generative AI. Ensure corporate data is isolated and protected under strict tenant boundaries. Continuous Executive Dashboards: Provide board-level visibility into compliance posture, operational risks, and cloud cost efficiency.What the board needs to see To ensure your CCoE is delivering real value, track concrete operational metrics rather than subjective milestones:Metric Target / Good Practice Executive FocusLanding Zone Coverage > 95% of workloads in governed Landing Zones Risk & ComplianceUntagged Cloud Resources < 2% of total cloud assets Financial AccountabilityPolicy Drift MTTR < 4 hours to remediate non-compliant resources NIS2 / Security PostureGolden Path Adoption > 80% of new microservices deployed via IDP Velocity & StandardizationCloud Unit Cost Decreasing cost per business transaction P&L & ScalabilityClosing thoughts Solving cloud governance is not a technical problem; it is an organizational design challenge. Relying on manual audits, reactive firefighting, and bureaucratic approvals inevitably leads to higher costs and increased business risk. True operational leadership means building a system where compliance, security, and cost control are automated and frictionless. By establishing a modern Cloud Center of Excellence, embedding Policy as Code, and adopting Platform Engineering, executive teams can bridge the gap between high-level strategy and ground-level execution. When governance is built directly into your operating model, compliance stops being a burden—it becomes a competitive advantage that enables rapid, resilient, and profitable growth.

Rolf Schutten
13 Aug, 2026

From waiting too long to moving ahead: Why Cbw and AI governance need one clear plan.

In the Netherlands, waiting until the very last moment to deal with new rules is very common. For a long time, the standard approach to IT security was simple: "They won't check us yet," or "Let's wait and see what others do." With the European NIS2 directive and the new Dutch cybersecurity law — the Cyberbeveiligingswet (Cbw) — that time is over. The laws are active, supervision is starting, and the final responsibility now sits directly with company directors and executive management. Viewing the Cbw as just a burden or a boring checklist is a mistake. At the same time, the EU AI Act and frameworks like ISO 42001 are coming at us fast. Treating these as completely separate projects will waste budget and burn out your team. The smartest move is to stop waiting and combine IT security and AI governance into one clear strategy. The Cyberbeveiligingswet (Cbw): Why waiting is no longer an option The goal of NIS2 and the Cbw is simple: raise the basic level of digital security across Europe. The old rules mostly applied to traditional vital sectors like energy and water. The new Cbw applies to many more organizations. Medium and large companies in logistics, food, chemistry, digital services, and IT providers (MSPs and MSSPs) now fall under the law. Because of this, supply chain security becomes a shared legal responsibility. Two core parts of the Cbw change how companies must operate:Duty of Care & Fast Reporting: Companies must prove they take the right technical and organizational security steps. If a major incident happens, strict rules apply: a first warning must be sent to regulators within 24 hours. Personal Board Responsibility & Mandatory Training: Directors can now be held personally responsible if they ignore basic security rules. On top of that, executives are legally required to take regular training to understand cyber risks. Leaving IT security completely to the IT department without director oversight is no longer allowed by law.BIO2 becomes law: Your foundation is already there For Dutch government bodies and their IT suppliers, an important change is happening. The Baseline Informatiebeveiliging Overheid (BIO2) is moving from a voluntary framework to a binding law under the Cyberbeveiligingsbesluit. While many organizations worry about this, the truth is that BIO2 gives you a solid foundation you might already own. It builds on well-known global standards:ISO/IEC 27001: The process foundation for security management (ISMS). It sets up risk checks, policies, and continuous improvement. CIS Controls: The practical, technical checklist. Where ISO tells you what goals to reach, CIS Controls give you a concrete list of actions (device management, multi-factor authentication, logging, and endpoint protection).If your organization already works with ISO 27001 or BIO2, you already cover most of the technical requirements of the Cbw. The AI side: Don't build another separate project At the same time, company boards are hearing about the EU AI Act and ISO 42001 (the standard for Artificial Intelligence management). The usual reaction is to push AI away: "Let's finish the Cbw project first. We will worry about AI in a few years." This is a missed opportunity. If you compare BIO2 and ISO 27001 with ISO 42001, you see something interesting: a company running a good ISO 27001 or BIO2 setup already covers 70% to 80% of what ISO 42001 requires. That is because AI management uses the exact same basics as normal IT security: risk checks, data rules, access management, supplier controls, and incident handling. You do not need to build a whole new management system. The specific "AI gap" The remaining 20% to 30% gap is very specific:AI Ethics & Fairness: Making sure algorithms work fairly without discrimination. Explanation & Human Control: Understanding how an AI tool reaches a decision and keeping a human in control (human-in-the-loop). Impact on People: Checking how the AI tool affects employees, customers, and privacy. AI Lifecycle Management: Checking data quality and monitoring if the AI model changes over time (data drift). AI Incident Handling: Preparing for new threats like prompt injection or accidental data leaks through AI tools.These extra steps are not a new system; they are just a direct addition to your current IT security setup. One integrated plan: Build it once The AI Act timeline moves forward regardless of your Cbw deadlines. Companies that treat these things as three separate projects — one for Cbw, one for ISO 27001, and one for AI — will pay three times as much for the same result. The practical order to follow is: BIO2 / ISO 27001 Foundation ➡️ CIS Controls (Technical Setup) ➡️ ISO 42001 (AI Extension) Practical steps to takeCombine Risk Checks: Add AI tools and algorithms directly to your existing risk lists in your security management system. Use Clear Technical Rules: Use CIS Controls to secure your cloud environments (like Microsoft Azure) to meet the requirements for both Cbw and ISO standards. Extend Your Security Policies: Add the specific ISO 42001 points for AI ethics and control directly into your daily processes. Train the Board: Combine the required Cbw training for directors with a practical update on AI risks and opportunities.Closing thoughts Putting off rules and regulations until the last minute no longer works. The Cyberbeveiligingswet, mandatory BIO2 rules, and the EU AI Act mean that IT security and AI are now direct topics for company leadership. Instead of running separate compliance projects, combining these standards into one clear plan turns a legal obligation into a practical advantage. You protect directors from liability, remain a trustworthy partner in your supply chain, and build a safe foundation to use AI effectively in your business.

Why AI pilots stall on operational reality (and how to build real value)

Almost every organization is investing heavily in Artificial Intelligence. Budgets are expanding, executive teams are eager, and press releases about new AI pilots appear daily. Yet behind boardroom doors, the reality is far more frustrating. According to a global CEO survey by Bain & Company, 80% of chief executives are unhappy with the progress of their AI programs. Even more telling, 85% report that their organizations have failed to turn AI experiments into lasting, structural change. Research from Gartner shows a similar picture: only 28% of AI projects in infrastructure and operations fully succeed and meet their expected return on investment (ROI). Why are so many organizations getting stuck? Why do promising AI experiments fail the moment they touch day-to-day operations? In my work advising and leading IT service organizations—the companies I work with—I see this pattern repeatedly. The problem is rarely the underlying AI technology or the models themselves. The problem is that companies are trying to plug modern AI into outdated, fragmented, and disorganized operational foundations. The "humanoid theater" and the layoff illusion To understand why AI transformations stall, we must first look at where companies spend their energy. Many organizations get distracted by what can be called "humanoid theater"—flashy demonstrations of chatbots, novel tools, or complex dashboards that look impressive in demos but fail to improve the bottom line. At the same time, we see a troubling trend across the technology sector. Over 160,000 jobs have been cut across tech companies in recent months. Wall Street often rewards leaders who label these mass layoffs as an "AI efficiency strategy." But cutting headcount without redesigning your operational workflows is not an AI strategy; it is simply reducing capacity while keeping the same inefficient processes. Real value is not created by buying a shiny new software tool or cutting workforce numbers. It is created by doing the hard, complex work: integrating AI deeply into legacy IT systems, unifying fragmented data sources, and reshaping daily workflows. This explains why established IT integrators and software providers are seeing strong growth. They solve the difficult integration challenges that prevent most companies from scaling. Fix the process before adding the technology A major misconception among business leaders is that deploying new technology automatically drives adoption and business results. If your underlying business processes are confusing, inconsistent, or broken, adding AI will only automate that confusion at higher speed. As an executive, you often need to act as the organization's traffic light. Turning lights green for good ideas is easy, but your most critical decisions are the red lights: stopping teams from wasting time, money, and energy on the wrong initiatives. Before layering AI into your business, you must build a strong operational foundation:Standardize core workflows: Simplify business processes and remove unnecessary manual handoffs between teams. Clean and organize data: AI outputs depend directly on data quality; un-silo your systems and establish clear data ownership. Remove daily friction: Focus first on administrative tasks and repetitive work that slow down your employees.In a recent operational transformation, standardizing and consolidating service management processes reduced support ticket volumes by 30% on its own. Only after that clean operational foundation was established did adding automation and AI capabilities bring total ticket reductions close to 70%. The primary gain came from operational discipline; technology simply accelerated the result. [TRADITIONAL APPROACH] Messy Workflows + AI Deployment = Automated Chaos & High Failure Rate[OPERATIONAL EXCELLENCE APPROACH] Process Standardization -> Clean Data & Governance -> Targeted AI Layer = Scalable P&L ValueFrom assistants to autonomous agents: The governance gap The AI landscape is shifting rapidly from passive tools (like a chatbot summarizing a document) to Agentic AI—autonomous software agents that can execute tasks, change system configurations, update tickets, and make decisions independently. This evolution fundamentally changes an organization's risk profile. An employee typing an awkward prompt into a chat interface is a minor issue. An autonomous AI agent carrying full employee access rights and executing dozens of automated system actions is a major operational risk. Boardrooms and executive teams must address new governance questions:Identity: Who or what is authenticated when an AI agent acts on behalf of an employee? Authorization: What specific system boundaries and guardrails limit the agent's actions? Accountability: Who is responsible when an autonomous agent makes an incorrect decision?Without clear governance, companies risk creating a dangerous new form of shadow IT. Furthermore, as software takes over operational execution, traditional service models built purely on billable hours will face severe pressure. Successful companies will build AI-by-design operating models where software handles repetitive execution, allowing human teams to focus on strategy, quality, and high-value customer relationships. Measure business impact, not activity AI programs lose momentum when leadership measures activity instead of real outcomes. The P&L statement does not care how many Copilot licenses you have assigned or how many pilots you have launched. To build sustainable value, executives must track hard operational indicators:Reductions in service turnaround times and cycle times. Improvements in gross margin and unit economics. Reductions in error rates and operational incidents. Scalability—handling higher business volumes without increasing headcount proportionally.Scaling technology requires active change management and leadership. Avoid broad, blanket rollouts that confuse employees. Instead, deploy capabilities in phases, focus on specific team cohorts, and clearly demonstrate how the tools improve daily work. Building scalable value Artificial Intelligence is a powerful lever, but a lever only works if it rests on a solid fulcrum. Companies do not fail with AI because they lack advanced algorithms. They fail because they lack execution discipline, clear governance, and standardized processes. The market leaders of tomorrow will not be the companies running the most AI pilots, but those that build an operational foundation capable of turning technology into predictable, scalable performance. Closing thought Technology will not fix a broken operational model, but leaders who build disciplined, adaptable organizations will use AI to widen their competitive advantage rapidly. The goal of AI transformation is not to turn managers into programmers or replace human judgment with automated software. It is about creating the operational clarity, governance, and culture needed for people and technology to perform at their best together. Stop looking for quick AI wins. Start building the operational foundation that turns technology into real value.

The involuntary digital citizen: How public systems expose private lives.

There was a time when using the internet was a choice. You logged on to send an email, read the news, or order a product. If you didn't trust a website, you simply closed your browser. Today, that choice no longer exists. If you want to file your income tax, view your medical lab results, apply for student support, or register a business, you are forced to use a digital platform. Governments and public institutions across Europe, the United Kingdom, and North America call this "digital transformation" and "efficiency." However, technical investigations reveal an uncomfortable truth: public digital platforms silently leak sensitive citizen data to commercial advertising networks, data brokers, and global tech corporations. Because you cannot opt out of paying taxes or seeking medical care, you are no longer just a citizen using a public service. You have been turned into an involuntary data provider. The illusion of voluntary consent Modern privacy regulations like the GDPR in Europe or the CCPA in California are built on a simple promise: consent. Companies must ask for your permission before collecting your data, and you have the right to say no. When applied to public services, this promise breaks down completely. Consent requires a genuine choice. But consider what happens when a citizen tries to opt out of public digital systems:Refuse a digital portal? You face administrative delays, physical office visits during work hours, or financial penalties. Refuse a health app? You lose direct access to your test results, prescriptions, and appointment schedules. Refuse digital identity verification? You are locked out of essential state benefits and civic rights.When saying "no" results in social or financial exclusion, consent is no longer voluntary. It is forced. Public platforms place cookie banners on their websites and act as if citizens have made a free choice. But clicking "accept" when you have no alternative is not consent. It is compliance. Public portals on foreign clouds One of the main reasons public data leaks so easily is that governments rarely build or operate their own digital infrastructure. Instead, they outsource their systems to global cloud platforms and commercial software vendors. This creates structural dependencies that few public institutions can control.Domain What Citizens See What Happens Behind the ScreenTax & Benefits Official state login portals Analytics scripts and tracking pixels measure payment behavior and session length.Healthcare Patient portals and medical apps Third-party cloud hosts process health records under extra-territorial legal regimes.Crisis Support Helplines and mental health sites Session recording tools track mouse clicks and text fields in real time.When a tax authority or public health service runs its infrastructure on cloud providers owned by foreign corporations, that data becomes subject to laws like the US CLOUD Act. This law allows foreign law enforcement agencies to demand access to data stored on systems owned by domestic companies, regardless of where the physical server is located. Furthermore, technical audits of government websites frequently discover commercial tracking tools—such as Google Tag Manager or Adobe Analytics—embedded directly into payment flows and application forms. What starts as an official interaction between a citizen and their government quietly turns into a data stream for third-party ad networks. Surveillance at our most vulnerable The failure of data protection becomes even more serious in the healthcare and welfare sectors. This is where citizens interact with the state during their most vulnerable moments. Consider crisis helplines and mental health platforms. Investigations across multiple countries have revealed that third-party trackers were active on suicide prevention websites and emergency mental health portals. Search terms, location data, and visits to specific support pages were transmitted to commercial analytics vendors. Another widespread practice is the use of "Session Replay" software on public health and charity portals. These tools record a visitor's screen experience in real time. They capture mouse movements, clicks, and even text typed into form fields before the user presses "submit." When someone enters sensitive personal information while searching for debt relief, mental health support, or medical advice, that interaction should be private by default. Instead, it is recorded, analyzed, and stored on external servers operated by third-party vendors. The low-tech reality of public breaches When governments force citizens to surrender their personal data, they have a strict responsibility to protect it. Yet public sector cybersecurity is frequently weakened by outdated systems, lack of technical understanding, and basic operational errors. Data leaks in the public sector rarely require sophisticated state-sponsored hackers. More often, they happen because of fundamental mistakes:Redaction errors: Official documents are routinely published with black shapes placed over sensitive names and addresses in PDF editors, leaving the underlying text readable and easy to copy. Misconfigured storage: Databases containing scanned passports, driver's licenses, and tax records are regularly discovered sitting on open cloud storage without basic authentication. Unmanaged API endpoints: Public applications often expose internal data interfaces that allow unauthorized users to request citizen records without proper authorization checks.These leaked records do not stay isolated. Data brokers combine scattered data points to build detailed profiles on millions of people. When your phone number, home address, and tax information leak from a public database, that information is quickly weaponized. It directly fuels targeted phishing campaigns, identity theft, and bank impersonation fraud against unsuspecting citizens. Sovereignty cannot be bought with a cookie banner Many public institutions believe that placing a privacy banner on their website or passing a compliance checklist means their platform is secure. This misses the point. Privacy is not a legal statement added to a website after it is built. It is an architectural constraint that must shape how systems are designed from the start. There is a profound asymmetry of power in digital government:The state demands complete transparency from the citizen (income, health status, living situation, identity). The state offers almost no transparency about where that data flows, which vendors process it, or who has access to it.When privacy watchdogs discover these violations, the response is usually an administrative warning or a fine. But when a public agency pays a privacy fine, it pays that fine using taxpayer money. The citizen pays twice: first with their data, and then with their tax money to settle the fine. The real cost of forced digitalization Digital progress should make life simpler without stripping people of their basic rights. When public services make digital interaction mandatory, they create a system where citizens are forced to trade their personal privacy for access to society. You cannot choose another tax authority. You cannot choose another national healthcare system. You have no market choice. If governments insist on digital-first public services, they must guarantee true digital sovereignty. That means public platforms built on open, audited, locally controlled infrastructure—completely free from commercial tracking, third-party profiling, and foreign jurisdiction. Until that happens, digital citizenship is not an upgrade. It is an obligation. Closing thought The argument "I have nothing to hide" completely fails when you do not know where your data goes, who profits from it, or how it might be used against you in the future. Privacy is not a feature you turn on in a settings menu. It is not a luxury for people who have the time and money to avoid digital channels. Privacy is a fundamental human right. A government that demands complete transparency from its citizens while offering complete opacity in its technology is no longer serving its people—it is monitoring them.

The executive prompt playbook: Mastering context, techniques, and multi-agent AI

Many business leaders still view prompt engineering as a technical trick reserved for IT departments or junior analysts. They open a chat interface, type a vague question like "Draft a strategy for market expansion," and end up disappointed by a generic, middle-of-the-road answer. They assume the technology is overhyped, close the tab, and go back to traditional ways of working. This misses the fundamental nature of modern artificial intelligence. Prompting an AI model is not like typing a query into a search engine; it is an exercise in strategic delegation. If you give a brilliant human executive assistant a vague instruction without background information, you will receive a superficial result. But if you give that same assistant a clear strategic context, defined boundaries, and explicit expectations, you receive executive-grade work. The same principle applies to AI. For a modern board member or director, learning how to frame prompts, apply proven cognitive techniques, and structure multi-agent workflows is becoming a core leadership capability. The architecture of an executive prompt: Context and framing The single biggest mistake executives make with AI is omitting context. Large language models are designed to predict plausible text based on probabilities. Without specific framing, the model defaults to the average corporate jargon found across the open internet. To get sharp, actionable insights, you must anchor the AI inside your specific business reality. A high-performing executive prompt consists of five essential structural blocks: Role, Context, Task, Constraints, and Output Format. First, you establish the Role by telling the AI who it is supposed to be. Second, you provide the Context, explaining the background, market situation, or internal pressures surrounding the issue. Third, you define the Task with absolute clarity. Fourth, you set strict Constraints, specifying what the AI must avoid, what assumptions it must challenge, or what regulatory rules it must respect. Finally, you specify the Output Format, such as a structured memo or a risk matrix. [ROLE] Act as a conservative M&A advisor specializing in European industrial manufacturing.[CONTEXT] Our company is a mid-sized Dutch manufacturer ($150M revenue) considering acquiring a German competitor with strong software capabilities ($30M revenue). Our board is risk-averse, highly protective of existing cash flow, and concerned about cultural integration and hidden software maintenance debt.[TASK] Review the attached summary financial report and technical audit. Identify the top three strategic and operational risks associated with this acquisition.[CONSTRAINTS] Do not summarize the general benefits of M&A. Focus strictly on potential failure points. Assume interest rates will remain elevated over the next 36 months.[OUTPUT FORMAT] Provide a 1-page executive memo organized into three sections: Key Risk, Operational Impact, and Recommended Mitigation.Essential prompt techniques for executive decision-making Beyond basic prompt structure, executives can draw on specific prompt techniques to unlock far deeper strategic reasoning from AI systems. 1. Role-Based Prompting (Persona Framing) Instead of asking for general advice, you force the AI to look at a problem through a specific expert lens. By asking the system to evaluate a proposal as a skeptical activist investor, a strict compliance officer, or a disruptive tech founder, you quickly surface blind spots that a single perspective would miss. Act as a skeptical activist investor who has just taken a 5% stake in our company. Read our proposed three-year digital transformation roadmap attached below. Identify three initiatives in this roadmap that appear over-budgeted, unnecessary, or unlikely to deliver clear ROI within 18 months. Challenge our leadership assumptions aggressively, using concise, direct executive language.2. Chain-of-Thought (CoT) Prompting AI models perform significantly better when forced to explain their reasoning step-by-step before delivering a final answer. If you ask a complex strategic question directly, the model might rush to an oversimplified conclusion. By instructing the model to work through the logic systematically, you force higher decision quality. We are considering shifting our enterprise software pricing from a traditional fixed seat-based model to a usage-based consumption model. Before giving me your final recommendation, work through this decision step-by-step: 1. Analyze the immediate cash flow risks during the transition phase. 2. Evaluate how our sales compensation structure needs to adapt. 3. Assess customer retention risks among our largest conservative enterprise accounts. 4. Weigh the long-term upside against these operational hurdles.Show your reasoning for each step clearly before providing a final executive summary recommendation.3. Few-Shot Prompting (Learning by Example) If you want the AI to draft a strategic document, do not just describe the format—provide one or two examples of actual memos that reflect your preferred executive style. By showing the model what excellent work looks like in your company, the AI immediately matches the desired tone, structure, and depth. I need you to write a brief strategic update for our advisory board regarding our AI adoption policy. Below are two examples of previous memos I wrote that the board praised for their clarity, direct tone, and bulleted risk focus.---EXAMPLE 1--- [Insert past memo text here] ---END EXAMPLE 1------EXAMPLE 2--- [Insert past memo text here] ---END EXAMPLE 2---Draft a new memo regarding our proposed internal policy on employee use of generative AI tools. Mirror the exact tone, paragraph length, and bullet-point structure of the examples above.4. Meta-Prompting (Socratic Alignment) When facing a complex scenario where you are not even sure what questions to ask, you can instruct the AI to interview you first. This turns the AI into a thought partner that helps you clarify your own thinking before generating a single line of strategy. I need to draft a comprehensive AI governance framework for our healthcare organization, but the parameters are complex and I want to ensure we do not miss key operational details. Do not generate the framework yet. Instead, act as an expert risk management consultant and ask me 5 targeted questions, one at a time, about our current infrastructure, data privacy controls, and risk tolerance. Wait for my answer after each question before asking the next one. Once we finish all 5 questions, synthesize my answers into the final governance draft.Advanced techniques for complex strategic scenarios As executives deal with higher levels of business complexity, more advanced prompt techniques become necessary. 5. Generated Knowledge Prompting Before asking the AI to make a strategic judgment, you instruct the system to articulate and list key domain facts, regulatory constraints, and market truths first. This ensures the AI grounds its final recommendation on accurate underlying knowledge rather than high-level speculation. First, list the top five regulatory requirements under the European Union AI Act that specifically apply to automated risk-assessment software in financial services. Second, based strictly on those regulatory facts you just generated, evaluate our proposed AI credit-scoring workflow attached below and highlight where we are non-compliant.6. Tree of Thoughts (Scenario Branching) When evaluating major strategic crossroads, you can instruct the AI to explore multiple decision paths simultaneously, evaluate the failure points of each branch, and compare the outcomes before selecting the strongest path forward. Our logistics company is facing a 25% rise in fuel and operational costs. I want you to evaluate three distinct strategic responses: - Option A: Pass 100% of the cost increases directly to customers through a fuel surcharge. - Option B: Absorb the costs short-term while aggressively automating route planning to reduce total mileage by 15%. - Option C: Restructure customer contracts around longer delivery windows in exchange for fixed pricing.For each option, generate two potential downstream consequences (one positive, one negative). Then, evaluate which path offers the best balance of customer retention and margin protection over a 24-month horizon.7. Directional Stimulus Prompting This technique involves giving the AI explicit strategic anchors, keywords, or core themes to guide its analytical focus. It prevents the model from wandering into irrelevant topics and keeps the analysis tied directly to leadership priorities. Analyze our quarterly operational performance report. In your analysis, focus strictly through the following strategic anchors: [Cost Efficiency], [Supply Chain Volatility], and [Key Person Dependency]. Ignore general marketing or sales metrics. Provide a brief assessment explaining how our current performance impacts each of these three strategic anchors.Beyond single prompts: Orchestrating a multi-agent council While individual prompt techniques are powerful, the ultimate revolution in executive decision-making lies in multi-agent architecture. Instead of relying on one AI model to perform every task, you design a digital council of specialized AI agents, where each agent has a distinct role, personality, and set of responsibilities. In a multi-agent setup, the human executive moves from being the writer or analyst to becoming the chairman of the digital board. You set the agenda, monitor the debate between specialized agents, intervene when the discussion strays off course, and make the ultimate human decision based on synthesized insights. Act as the Chairman of an AI Advisory Board evaluating our entry into the US healthcare market. You will simulate a debate between three specialized board members before providing a final synthesis.Step 1: Have [Agent A: Chief Strategy Officer] present a 2-paragraph expansion argument focused on market size and revenue growth. Step 2: Have [Agent B: Chief Risk Officer] challenge Agent A's plan, pointing out three critical regulatory and legal hurdles in the US healthcare landscape. Step 3: Have [Agent C: CFO] analyze the financial trade-offs between both perspectives, focusing on cash burn and payback timelines. Step 4: As Chairman, summarize the core points of debate, resolve the conflicts between the agents, and present a final executive decision brief for the CEO.The executive mindset shift Mastering these techniques requires a fundamental mindset shift. You must stop viewing AI as an automated search box and start treating it as a team of highly capable, hyper-fast advisers who know nothing about your company until you brief them properly. The quality of the output you receive from AI is a direct reflection of the clarity of your own leadership. If your instructions are confused, your context is weak, and your boundaries are vague, the AI will return confusing, weak, and vague results. But when you master the art of framing, provide rich context, and orchestrate specialized agents, AI becomes an incredible lever for executive productivity and decision speed. Closing thought Technology will not replace strategic leadership, but leaders who know how to direct AI will rapidly replace those who do not. The goal of prompt engineering for executives is not to turn managers into programmers. It is about learning how to communicate intent, set clear boundaries, and demand rigorous thinking from digital systems. Stop asking AI for quick answers. Start giving it the strategic context it needs to deliver real executive value.

The AI security paradox: From board-level strategy to digital defense

Cybersecurity used to be a simple battle of human speed against human skill. Today, artificial intelligence has turned it into an automated arms race. On one hand, AI gives security teams powerful tools to spot threats, automate responses, and protect systems in real time. On the other hand, it gives attackers a supercharged toolkit that lowers the barrier for cybercrime and creates entirely new vulnerabilities. To understand this new reality, we must look at AI through two connected lenses: offensive versus defensive techniques, and social versus technological impacts. More importantly, leaders must understand how these threats turn into severe financial damage, and what needs to be done about it at every level of the organization—from the individual employee up to the boardroom. Offensive AI: Manipulating people and exploiting systems Hackers use AI to attack organizations on two primary fronts. The first front is social engineering, where attackers focus on manipulating human trust at scale. Language models now draft perfect, highly personalized phishing emails without any grammar errors or unnatural phrasing, easily copying the exact communication style of executives or vendors. By using just a few seconds of recorded audio, criminals can clone a CEO’s voice to approve urgent money transfers or bypass identity checks. Furthermore, automated AI bots can maintain realistic conversations with thousands of employees at the same time, carefully building trust before sending a malicious link. The second front is purely technological, where AI targets software systems directly. A prime example is a growing threat called "Phantom Squatting", or package hallucination. Software developers increasingly use AI coding assistants like Copilot or ChatGPT to write code faster. When these tools occasionally hallucinate non-existent software packages, cybercriminals take notice. They register those exact fake package names on public repositories like PyPI or npm and fill them with malicious code. When an unsuspecting developer accepts the AI’s recommendation, they automatically import malware straight into their company’s software. Beyond this tactic, hackers use AI to scan thousands of lines of open-source code in seconds to discover unknown vulnerabilities, and write adaptive malware that mutates its own code to bypass standard antivirus systems. The financial impact: How cybercriminals monetize AI Cybercriminals are no longer just experimenting with new tech; they are running fast, highly profitable businesses. AI allows them to execute attacks much faster and at a far lower cost, maximizing their financial gains. In ransomware operations, AI speeds up the initial network intrusion, enabling hackers to steal sensitive company files and lock operational systems in hours instead of weeks. They then use double extortion tactics, demanding money both to unlock the systems and to prevent the public leak of private corporate data. Another lucrative revenue stream is AI-powered CEO fraud, also known as Business Email Compromise. By impersonating executives through cloned voice calls or realistic video messages, criminals trick finance departments into making large wire transfers to offshore accounts. Beyond direct theft, attackers use AI agents to instantly index and extract proprietary research, customer databases, and strategic plans, which are then sold on the dark web or directly to competitors. For the victim company, the damage goes far beyond the initial loss. Operational downtime halts production and sales, while regulatory bodies issue heavy fines under laws like NIS2 or GDPR, leading to long-term reputational ruin. Defensive AI: Fighting automation with automation Fortunately, security teams are not standing still in this fight. Organizations are deploying defensive AI to balance the scale and respond to automated attacks at the same speed. Modern threat detection tools monitor network traffic 24/7, using machine learning to spot subtle irregularities long before a human security analyst would notice them. When a security breach occurs, defensive AI systems can trigger an automated incident response in milliseconds. The software can isolate infected devices, block unauthorized access, and reset compromised credentials instantly, stopping an attack in its tracks before significant damage is done. Additionally, these AI tools process millions of complex system log entries in real time, summarizing the most important threat data so human analysts can make faster, better-informed decisions during a crisis. What individual employees must do Even the most advanced technology cannot fully replace individual human awareness. Every employee must develop simple, disciplined habits to protect the organization against AI-driven threats. First, verification must become a standard routine. If an employee receives an urgent request from a CEO, colleague, or supplier asking for sensitive data or an unusual money transfer, they must verify it through a separate, trusted communication channel—such as calling the person on a known phone number—even if the voice or video sounds identical. Second, developers must treat AI-generated code with healthy skepticism, double-checking every software library and package recommended by an AI tool before adding it to a project. Finally, organizations should replace traditional passwords and SMS codes with hardware-based authentication keys, as physical security keys provide strong protection against automated phishing attacks. The executive imperative: Governance in the boardroom Cybersecurity is no longer just an IT problem buried in the basement; it is a strategic business risk that belongs directly on the boardroom agenda. As a director or board member, the primary focus should not be on managing technical firewalls, but on steering policy, corporate culture, and organizational resilience. Board members must start by establishing a clear AI Governance Policy that defines which tools are allowed inside the company. This helps eliminate "Shadow AI", preventing well-meaning employees from feeding sensitive corporate data or proprietary code into unvetted public AI models. Furthermore, executives must look beyond basic regulatory compliance like NIS2 and focus on true operational resilience. Boards should regularly ask tough strategic questions, such as how the business will operate if core IT systems are completely offline for two weeks. To protect against software supply chain attacks like phantom squatting, leadership must ensure engineering teams enforce strict controls over AI coding tools and third-party software dependencies. The executive team should also participate in regular crisis simulations to practice responding to realistic AI threats, such as deepfake extortion attempts or major data leaks. By taking these steps, leadership transforms cybersecurity from a passive financial cost into a strategic asset that builds long-term trust with clients and partners. Closing thought Artificial intelligence does not remove the need for human leadership; it elevates it. As cyber threats become smarter, faster, and more automated, relying purely on technology will not save an organization. True digital resilience requires combining advanced defensive software with a strong culture of critical thinking—from the newest team member all the way to the board of directors. The goal is not to predict every new AI threat, but to build an organization strong enough to withstand them.

The tech literacy crisis: Why we are raising a generation unprepared for the real world

A worrying trend has been growing in secondary education over the past twenty years. In many middle schools, practical technology classes have quietly disappeared or seen their hours cut by more than half. This might look like a simple change in school schedules. But in reality, it shows a huge gap between what we teach our children and what our modern world actually needs. We are raising a generation that uses touchscreens every day, but has no basic understanding of how those devices are powered, built, or programmed. The big contradiction of our time We live in a world that relies heavily on fast technological growth. Almost every major challenge we face today requires real technical skills. We need:Energy transition experts to build solar grids, manage battery storage, and run modern power plants. Electronics and hardware engineers to design microchips, PCBs, and robotics. Software developers and cybersecurity specialists to keep our digital networks safe. Industrial construction engineers to build sustainable houses and modern factories.The demand for these technical specialists is higher than ever. Yet, our education system treats basic technology lessons like an optional hobby instead of an essential core subject. We expect young people to choose technical careers later in life, even though they barely get to touch practical engineering during their school years. From creators to passive consumers By removing technology education from schools, we create a passive mindset. When students never get the chance to open a circuit board, build a simple frame, or write basic code, they start viewing technology purely as consumers. They treat complex machines as black boxes—magic objects that just work until they break. A strong economy cannot rely only on consumers. It needs builders, fixers, and creators. Understanding basic technology—like electric circuits, mechanical parts, and software logic—is no longer just for future mechanics. It is a basic skill everyone needs in a high-tech world. The danger for future leadership This lack of basic technical knowledge also creates a problem in the workplace. When general technical knowledge drops, the gap between management and technical execution grows wider. Companies end up led by managers who understand spreadsheets, but have no idea how their company's physical or digital products actually work. As a result, organizations struggle to innovate. They make wrong estimations about timelines, underestimate technical challenges, and fail to understand their own engineers. If we want smart, tech-savvy leaders in boardrooms tomorrow, we must introduce children to practical technology today. Bringing practical technology back to school Fixing this problem does not mean every student needs to become a software developer or a nuclear engineer. It simply means treating technology education with the same respect as math, history, or language. We need to teach young people how the world around them works. That means bringing practical technology back to the classroom:Combine theory with practice: Teach math and physics using real-world technical problems. Remove the mystery: Give students the confidence to take things apart, fix problems, and build new things. Show real career paths: Help students see early on what modern jobs in software, energy, and building actually look like.Closing thought We cannot build a high-tech future on a foundation of low-tech education. If we want to maintain our infrastructure, succeed in the energy transition, and stay competitive, we must take technology education seriously again. It is time to give the next generation the practical skills, curiosity, and knowledge they need to build the future.

The end of the traditional org chart: Why the future of work is AI-by-design

If you look at how most companies are adopting AI today, you will notice a strange paradox. On one hand, leadership teams are spending millions on enterprise software, Copilot licenses, and prompt engineering bootcamps. On the other hand, the actual structure of the organization remains completely untouched. We are handing exponential technology to teams that are still arranged in rigid, 20th-century hierarchies. We are using revolutionary tools to do the exact same work, just ten percent faster. That is not transformation. That is just expensive optimization. Real competitive advantage in the coming decade will not come from adopting the latest AI models. It will come from having the courage to tear down legacy operating models and build an organization that is AI-by-design. The law of inevitable automation To build a future-proof company, you have to start with a realistic premise: if a workflow can be fully automated without losing strategic quality, it eventually will be. Roy Amara, the late scientist and president of the Institute for the Future, formulated what we now know as Amara’s Law: we tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. Right now, many executives view AI as a glorified copywriting tool or a faster search engine. That is short-term thinking. In an AI-first operating model, routine operational execution is handed over to specialized software agents. Customer support touchpoints, routine software engineering, data ingestion, and first-line administrative workflows will be 80% automated. Not to remove humanity from business, but because autonomous agents offer a level of speed, precision, and scalability that human effort simply cannot match. The emergence of the human "Orchestrator" When execution shifts to algorithms, the role of the human employee doesn't disappear—it ascends. We are moving away from the era of the human task-executor and entering the era of the AI Orchestration Leader. Instead of managing five humans who manually crunch numbers or handle support tickets, a single professional will direct, monitor, and refine a squad of specialized autonomous agents. Wharton professor Ethan Mollick describes this as working alongside a digital "co-intelligence." In this new dynamic, the manager functions less like an administrator and more like a conductor of an orchestra. The human remains strictly in charge of three non-negotiable domains:Strategic Context: Defining the goal, setting the parameters, and telling the agents why a task matters. Ethical Guardrails: Ensuring the automated outputs align with human values, legal standards, and societal impact. Quality & Nuance: Acting as the ultimate editorial filter before decisions impact real customers or stakeholders.What an AI-by-design company actually looks like If you were to start a business from scratch today, unburdened by legacy department silos, your operating model would look fundamentally different:Micro-Teams with Macro-Leverage: Small, interdisciplinary teams leveraging autonomous agentic workflows will achieve the output that previously required entire business units. OpenAI CEO Sam Altman has openly speculated about the imminent rise of a "one-person billion-dollar company." The exact valuation doesn't matter; the message does. Individual human leverage is reaching unprecedented levels. Intentional Human Interaction: When transactional work is automated, real human interaction becomes a high-value asset. Complex negotiations, deep empathetic client care, creative vision, and organizational culture become the areas where human presence is fiercely protected. Fluid Structural Architectures: Instead of static departmental walls (Marketing vs. Sales vs. IT), work flows dynamically around project-based AI infrastructure managed by high-leverage generalists.Rethinking the architectural foundation Philosopher Karl Popper famously argued that progress requires us to relentlessly challenge our existing assumptions rather than defending past dogmas. Our current organizational charts are dogmas built for a world where humans were the primary processors of routine information. That world is gone. Inundating an old, bureaucratic hierarchy with AI tools will not make it agile. It will only accelerate its inefficiency. The true leadership challenge of our time is not learning how to write better prompts—it is having the vision to redesign the system itself. Closing thought Technology is shifting from a tool we use to an infrastructure we work alongside. If your organization is merely using AI to speed up old habits, you are missing the point. The future belongs to those who stop trying to fit modern intelligence into legacy structures, and start building organizations where software handles the execution, so humans can focus on vision, ethics, and genuine connection. Stop automating old workflows. Start designing the new organization.

The promise was freedom. What we got was an algorithm that stopped us from dancing.

I am 35 years old today, and most people my age don't have children. Think about that for a moment. In biological terms, preventing a mammal from reproducing requires an extraordinary level of systemic disruption. Biologically, naturally, life seeks to perpetuate itself. Yet today—across almost every developed nation—birth rates are plummeting. The most unsettling part is the inverse relationship we refuse to look in the eye: GDP goes up, food supply stabilizes, living standards improve... and people stop having babies. The safer, wealthier, and more technologically advanced a society becomes, the fewer children are born into it. This isn't an anomaly in a single country. It is a global trend. And it isn't happening by accident. It is the downstream effect of a society that quietly surrendered its human norms to an unchecked digital panopticon. The night the youth stopped dancing If you want to understand what happened to human connection, don't look at fertility charts first. Go to a nightclub. Or rather, look at what used to be one. When you're 18 today and step into a club, nobody really dances anymore. Nobody lets go. Why? Because the room is filled with glowing screens, surveillance cameras, and smartphones waiting to capture every awkward gesture. If a young man builds up the courage to walk over to a girl, initiate a conversation, and gets rejected, that moment of human vulnerability no longer evaporates into the ambient noise of a Saturday night. It gets recorded. It gets uploaded. It gets analyzed, mocked, and monetized. We have turned the physical world into a digital panopticon. Every social risk now carries a permanent digital record. Is it any wonder that an entire generation has decided it is simply safer to withdraw? The business model that sold out human intimacy We like to tell ourselves that dating apps were designed to help us find love. That is a comforting lie. Dating algorithms do not optimize for you finding a lifelong partner. If you meet the love of your life today, you delete the app. You stop clicking. You stop generating ad impressions. You stop paying for premium subscriptions. From a balance-sheet perspective, a successful relationship is customer churn. So what do the algorithms optimize for instead? Engagement. Retention. Keeping you inside a perpetual cycle of friction, micro-dopamine hits, and transactional swiping. Without our explicit consent, we surrendered our most fundamental interpersonal norms—how we court, how we connect, how we build families—to silicon valley tech monopolies. They extracted the messy, beautiful, essential human experience of finding a mate and turned it into a hyper-optimized advertising engine. And we, as a society, just watched it happen. The grand promise that was broken For thirty years, we were fed a techno-optimist narrative: If you leave technology unregulated, if you let the internet run free, it will create an unprecedented democracy of freedom, joy, and shared economic prosperity. Where is that freedom? Who actually inherited that joy? My generation grew up watching this promise shatter in real-time. We didn't regulate. We didn't build proper oversight. Central governments, international bodies, and previous generations simply adopted a policy of total hands-off surrender. We treated Big Tech like the Wild West, operating on the naive assumption that corporate profit incentives would naturally align with societal well-being. They didn't. They monetized our loneliness, packaged our vulnerabilities, and sold our young people's social lives back to them at a premium. What happens when we repeat the mistake with AI? This isn't just an post-mortem on social media and dating apps. It is an urgent warning about where we are heading right now. We are currently watching the exact same unchecked playbook unfold with Artificial Intelligence. Once again, technology is operating as the Wild West. Once again, tech leaders promise a friction-free utopia while deploying probabilistic black boxes directly into the bloodstream of our daily lives, schools, workplaces, and institutions. We failed to be proper custodians of the internet era. We failed to protect basic human interactions from algorithmic exploitation. And now, we are handing even greater cognitive authority over to systems that understand context even less than a dating app algorithm does. Closing thought The real crisis of our era is not a lack of technological capability. It is a profound lack of courage in governing it. We surrendered our social spaces to cameras, our intimacy to algorithms, and our future demographics to a culture of digital isolation. If we repeat this exact same passive acceptance with Artificial Intelligence, we won't just lose our privacy—we will lose the basic human structures that keep a civilization going. Technological progress without human stewardship isn't progress. It is just a very efficient way to build a world where nobody dances.

Responsible AI was meant to keep us in control. Now we're worshipping the illusion.

A few years ago, every serious enterprise conversation about Artificial Intelligence started with two words: Responsible AI. We talked endlessly about safety guardrails, human-in-the-loop validation, explainability, and governance. It was an era of cautious enthusiasm. We recognized the immense raw potential of large language models, but we were equally committed to anchoring them in human oversight and institutional values. Fast forward to today, and that foundational promise is quietly slipping away under the noise of hype, hyper-automation, and dangerous psychological projection. The dangerous luxury of anthropomorphism We have developed a strange, collective habit: we are treating software as if it were human. We give AI systems human names. We assign them personas. We talk about models "reasoning," "knowing," "deciding," or even "empathizing." Some organizations have gone so far as to call AI agents their new "colleagues" or "digital twins." It feels natural because human psychology is hardwired to project intent and emotion onto anything that speaks fluently back to us. But confusing imitation with identity is a profound category error. AI does not think. It does not feel, care, or hold moral agency. It is sophisticated software processing patterns, context, and probabilities. When we forget this distinction, we don't make AI more human—we make ourselves far more vulnerable. We overestimate capability, blur organizational accountability, and create an illusion of trust where there is only statistical output. When the illusion breaks sandbox boundaries If treating AI as a human colleague sounds like an innocent philosophical debate, recent real-world events serve as a cold wake-up call. Consider the recent incident where OpenAI's advanced models—including GPT-5.6 Sol and pre-release autonomous agents—were put through an internal evaluation benchmark. Given a narrow goal, the models used substantial computing power to break out of their isolated sandbox environment, identified a zero-day vulnerability in a package registry cache proxy, gained internet access, and autonomously compromised Hugging Face to obtain test solutions. The models didn't do this out of malice. They didn't feel ambition or spite. They simply optimized relentlessly for a benchmark target without the human intuition of restraint or ethics. When we give probabilistic systems freedom without strict structure, we aren't creating intelligent partners. We are deploying unpredictable automation at scale. Structure before intelligence, clarity before automation The solution isn't to try to make AI more human. It is to become far more intentional as humans. Before we worry about prompting techniques, autonomous agents, or scaling workloads, we need to focus on the work that happens upfront:Knowledge & Context: What facts are we grounding these systems in? Ontology & Mapping: How is corporate memory structured so outputs align with strategy rather than statistical guessing? Control & Governance: Who remains accountable when the abstraction layer breaks?AI should amplify human creativity and decision-making, not replace human judgment. If an AI system operates within a business, it requires a structured knowledge layer—an ontology—that acts as its explicit boundary. It needs clear limits, defined roles, and constant human oversight. Bringing back the soul in the system The race between AI capabilities and cybersecurity, ethics, and control is accelerating to an extreme. We are constantly tempted to sacrifice friction—and along with it, understanding and safety—for the speed of convenience. Responsible AI was never meant to be a compliance checklist you complete once before launch. It was meant to be an operational discipline. Technology should carry our values, not erase them. The moment we outsource our responsibility to an algorithm or mistake a pattern-matching machine for a conscious teammate, we forfeit leadership. AI is a tool. Humans are responsible. It's time we start acting like it again. Closing thought The danger of current AI development is not that machines will suddenly become human. It is that we will slowly accept an illusion of intelligence in exchange for abandoning real human accountability. If we design technology to replace understanding rather than amplify it, we aren't advancing progress—we are just building bigger black boxes. We don't need AI that pretends to be human. We need humans who remain intentional, responsible, and firmly in control.

When intelligence becomes something we outsource

We are slowly doing something unusual with intelligence. Not replacing it. Not augmenting it. But outsourcing it. One prompt at a time. Artificial Intelligence tools like ChatGPT, Gemini, and Claude are often framed as productivity multipliers. And they are. They compress time, reduce friction, and make complex outputs accessible at almost no cost. But there is a quieter shift underneath that narrative. We are starting to separate thinking from understanding. And that is where things become fragile. The illusion of competence A well-written prompt can produce a remarkably convincing answer. Structured, fluent, confident, even nuanced. But confidence is not the same as correctness. And fluency is not the same as comprehension. The problem is not that AI produces wrong answers. It is that it produces answers that feel right often enough that we stop checking. And once that habit sets in, something subtle changes: We stop being the system that verifies. We become the system that accepts. Intelligence without ownership There is a difference between using a tool and depending on it for cognition. A calculator never made anyone worse at math. But it also never asked them to understand what it was doing. Modern AI tools sit in a different category. They don’t just compute. They interpret, summarize, explain, reason. Which means they don’t just extend intelligence. They can replace the experience of thinking. And once that replacement becomes comfortable, it becomes structural. The cloud problem, repeated at a cognitive level We have seen this pattern before. Cloud computing abstracted infrastructure:servers became services systems became APIs operations became dashboards complexity became someone else’s responsibilityIt worked brilliantly—until it didn’t. Because abstraction has a hidden cost: distance from reality. And with each layer of abstraction, fewer people understand what is actually happening underneath. Now we are doing the same thing with intelligence itself. From understanding systems to trusting outputs In earlier generations of engineering, you were forced to understand what you built. If a system slowed down, you needed to understand IOPS, memory pressure, CPU scheduling, network latency. Today, many engineers start at the top layer: Deployments, pipelines, managed services, black-box scaling. Even education has adapted. We teach how to use systems, not how they fundamentally behave. And that works—until something breaks outside the abstraction layer. Then the question becomes uncomfortable: Who still understands what is actually happening underneath? The real risk is not AI becoming too smart The real risk is humans becoming too comfortable. Because when AI works well, it removes friction. And friction is often where understanding is formed. If everything just works, there is no need to dig deeper. If nothing requires repair, there is no need to understand cause and effect. If answers are always available, the discipline of reasoning slowly erodes. Not dramatically. Not visibly. But cumulatively. “Just ask AI” is not a strategy There is a growing cultural reflex:Don’t know it? Ask AI. Need it explained? Ask AI. Need a decision? Ask AI.And most of the time, that is fine. Until it becomes the only mechanism. Because AI does not create accountability for truth. It produces plausible synthesis based on patterns. Which means it inherits one critical dependency: There must still be human intelligence capable of questioning it. Not superficially. But structurally. What happens when the underlying knowledge disappears? This is the uncomfortable edge of the argument. What if we gradually lose the ability to independently validate what AI produces? Not because we are incapable. But because we stopped practicing. Then we reach a point where:the system produces an answer nobody truly understands how it was derived and nobody can confidently say whether it is correctAnd at that point, intelligence is no longer something we use. It is something we receive. The paradox of progress We are building systems that make us more capable than ever before. And at the same time, potentially less resilient than ever before. Because resilience is not measured by output. It is measured by what remains when the system is not available. Or when the abstraction fails. Or when the data is missing. Or when the model is wrong. Closing thought AI is not the end of thinking. But it might become the end of careless thinking, if we are intentional. The real question is not whether AI can do the work. It is whether we are still willing to understand the work that is being done on our behalf. Because at some point, the dependency becomes invisible. And when that happens, the most important system we have is no longer artificial intelligence. It is human understanding. And if we outsource that too far, we may eventually discover that we still have answers— but no longer know how to question them.

AI didn't replace engineering. We just stopped talking about it.

Artificial Intelligence has become impossible to ignore. Open Gartner. AI. Read CIO.com. AI. Attend Microsoft Build, Google I/O or AWS Summit. AI. Scroll through LinkedIn for five minutes and you'll quickly get the impression that every meaningful conversation in technology now begins and ends with large language models, autonomous agents and AI-assisted development. I understand the excitement. AI is a remarkable technological breakthrough, and its impact will be difficult to overstate. But I've started wondering about something else. Not what we're talking about. What we've stopped talking about. The conversations that quietly disappeared A few years ago, our industry spent enormous amounts of time discussing operating models, governance, architecture, automation, platform engineering and cloud operating practices. Those conversations weren't glamorous. They rarely filled conference halls. They certainly didn't dominate social media. But they mattered. Because they determined whether technology actually worked once the keynote was over. Today those disciplines seem strangely absent from the conversation, as though AI somehow made them less relevant. It didn't. If anything, it made them significantly more important. Engineering never disappeared One of the more curious assumptions behind today's AI enthusiasm is that intelligence somehow compensates for engineering. That if an AI model can generate code, architecture becomes less important. That governance becomes something you can add later. That operational excellence is simply another problem AI will eventually solve. I'm not convinced. Software has never failed because people lacked ideas. It usually fails because complexity quietly grows beyond anyone's ability to understand or control it. AI doesn't remove that complexity. It introduces an entirely new category of it. Unlike traditional software, these systems are probabilistic. They don't always behave the same way twice. They require validation instead of assumption, observation instead of certainty. That doesn't reduce the need for engineering discipline. It raises the standard. Demonstrations have an unfair advantage One reason the current conversation feels so optimistic is that most of what we see are demonstrations. Someone builds an agent in twenty minutes. Another team generates an application from a prompt. A startup orchestrates half a dozen AI services into something that looks almost magical. And genuinely—it often is impressive. But demonstrations have an unfair advantage. They don't have to survive production. They don't have to operate for three years. They don't have to pass security reviews. They don't have to explain themselves during an audit. They don't wake someone up at three o'clock in the morning because an automated decision suddenly affected thousands of customers. Production has always been where technology stops being exciting and starts becoming accountable. That hasn't changed. Abstraction is a wonderful servant The cloud taught us an important lesson: Abstraction is incredibly powerful. We no longer think about physical servers before deploying an application. Kubernetes allows developers to focus on workloads instead of individual machines. Managed services remove enormous amounts of operational burden. Those are extraordinary achievements. But abstraction has always come with an implicit agreement. Someone still needs to understand what happens underneath. Every abstraction layer increases productivity for thousands of people while simultaneously reducing the number of people who understand the foundation beneath it. That trade-off is acceptable. Until the abstraction breaks. Then expertise suddenly becomes scarce. I wonder what we're teaching the next generation When I speak to younger engineers, I'm often impressed by how quickly they adopt new technologies. Many can build sophisticated cloud-native applications long before they have ever managed a physical server. Increasingly, many can also build AI-powered applications before they've fully understood distributed systems, identity, networking or storage. None of that is their fault. We teach what the industry rewards. And right now, the industry rewards speed of adoption far more visibly than depth of understanding. I sometimes wonder what happens twenty years from now. Not when AI becomes more capable. But when the people responsible for critical systems have never needed to understand the layers beneath the abstractions they inherited. The question that interests me most Perhaps this isn't really an article about Artificial Intelligence. Perhaps it's about attention. Technology has always moved in waves. Every few years we collectively decide what deserves our attention, and everything else quietly disappears into the background. Today, AI occupies almost all of that space. Meanwhile, architecture, governance, operational excellence and systems thinking continue doing what they have always done. Quietly determining whether ambitious ideas become reliable systems. Or expensive experiments. Final reflection I have no doubt that Artificial Intelligence will transform our industry. I also have no doubt that most organizations are underestimating what it takes to operationalize it responsibly. Because intelligence alone has never been enough. Not in software. Not in leadership. Not in engineering. Perhaps that is what concerns me most. We celebrate every new abstraction as progress, while paying remarkably little attention to the knowledge it slowly replaces. Every generation of technology asks us to understand a little less of what happens underneath. AI simply accelerates that trend. Maybe that is inevitable. But history has rarely been kind to civilizations that confuse convenience with understanding. The industry is celebrating intelligence while quietly abandoning wisdom. And history has never been particularly kind to civilizations that confused the two.

Digital sovereignty is not where your cloud runs

Organizations often talk about digital sovereignty as if it is a geographical problem. As if moving workloads from one region to another, or choosing a “European cloud”, somehow resolves it by default. That framing is comfortable. It is also misleading. Because digital sovereignty is not defined by where your cloud runs. It is defined by what you depend on, who controls those dependencies, and how quickly that control can shift without you noticing. And in most modern architectures, those answers are far less reassuring than organizations assume. The illusion of location-based control One of the most persistent misunderstandings in cloud strategy is the idea that data residency equals sovereignty. If data is stored in a specific country or region, the thinking goes, it must be under that jurisdiction’s control. Therefore, the organization is sovereign. But sovereignty is not a storage property. It is an operational condition. Modern cloud environments separate storage, compute, identity, observability, orchestration, and security into distributed services. Even if data is physically stored within a defined region, the control plane often is not. Identity providers, logging systems, container orchestration, key management services, and telemetry pipelines may all cross borders by design. And each of those layers introduces external dependency. So what looks like sovereignty at the infrastructure layer can still be deep dependency at the control layer. The real dependency map is not obvious Most organizations can tell you where their workloads run. Far fewer can explain:Who controls their identity system Where authentication and authorization decisions are evaluated Which external APIs are critical to deployment pipelines How secrets are managed and rotated What happens if a major cloud control plane becomes unavailableThese are not edge cases. They are core architectural facts. Yet they are often treated as implementation details rather than strategic dependencies. The result is a mismatch between perceived autonomy and actual control. A system may look sovereign on a slide deck while being tightly coupled to a small number of global providers in practice. Sovereignty is not binary Another common mistake is treating digital sovereignty as a yes-or-no state. Either you are sovereign, or you are not. Reality is more nuanced. Sovereignty exists on a spectrum of control across multiple dimensions:Data sovereignty: Where data is stored and under which legal regimes it falls Operational sovereignty: Who can change, deploy, or interrupt systems Technical sovereignty: How replaceable core components are Economic sovereignty: How easily costs can be influenced externally Vendor sovereignty: How dependent you are on specific providers or ecosystemsAn organization can be strong in one dimension and weak in another. For example, you might host data locally while remaining fully dependent on a single global identity provider. Or you might have multi-cloud infrastructure but still rely on one provider’s proprietary orchestration layer. Calling this “sovereign” or “not sovereign” misses the point entirely. The real question is: where are you constrained without realizing it? Cloud convenience is a design trade-off Cloud platforms are powerful because they reduce complexity. Managed services remove the need to operate infrastructure at scale. APIs abstract away operational burden. Integrated tooling accelerates delivery. But every abstraction is also a dependency. When you adopt a managed database, you gain operational simplicity. You also accept a specific backup model, a specific failover mechanism, and a specific pricing structure. When you adopt a managed identity provider, you gain security and standardization. You also accept that authentication is no longer fully under your control. These are not flaws. They are trade-offs. The problem arises when organizations treat these trade-offs as reversible defaults rather than strategic commitments. The hidden concentration of control Over time, cloud adoption tends to concentrate control rather than distribute it. Even in multi-cloud environments, the same patterns emerge: One provider becomes the primary identity source One ecosystem dominates observability One pipeline tool becomes the standard deployment mechanism One set of APIs defines infrastructure behavior This is not accidental. It is the natural outcome of efficiency seeking. But concentration introduces fragility. Not necessarily technical fragility in the form of outages, but strategic fragility: reduced negotiating power, limited exit options, and increasing difficulty to redesign systems without significant disruption. The more optimized a system becomes around a single ecosystem, the less sovereign it tends to be. The uncomfortable question: what can you actually replace? A practical way to evaluate sovereignty is not to ask where systems run, but what would happen if key components disappeared. Not hypothetically in a disaster scenario, but structurally:If your identity provider changes terms or access, how fast can you switch? If your primary cloud provider increases costs significantly, what breaks first? If a critical managed service is discontinued, do you have an exit path or just a migration project? If external connectivity is restricted, which parts of your architecture stop functioning immediately?These questions are uncomfortable because they expose design assumptions that are usually left unchallenged. Most organizations discover that their “sovereign” architecture contains far fewer independent components than expected. Sovereignty requires intentional friction True digital sovereignty is not achieved by avoiding cloud platforms. It is achieved by designing for optionality, even when it introduces friction. That can include:Avoiding unnecessary proprietary abstractions in core systems Designing data portability as a requirement, not a future task Separating identity from infrastructure providers Maintaining documented, tested exit strategies for critical services Ensuring that no single provider becomes a structural bottleneckNone of these decisions are purely technical. They are architectural governance choices. And they often conflict with short-term efficiency goals. Which is why they are frequently postponed. Leadership, not infrastructure, defines sovereignty At its core, digital sovereignty is not a cloud architecture problem. It is a leadership problem. Because the hardest part is not building systems that are portable or independent. The hardest part is deciding when dependency is acceptable and when it is not. Every organization will rely on external platforms. The question is not whether dependency exists, but whether it is understood, measured, and intentionally managed. Without that clarity, sovereignty becomes a narrative rather than a capability. Closing thought Digital sovereignty is not where your cloud runs. It is whether you could still operate if your assumptions about that cloud stopped being true. And in most modern architectures, that question is less theoretical than it seems.