Category
Articles about AI governance, responsible AI use, and organizational implementation of AI regulations.
50 articles
No framework prescribes an agent register with fixed fields, but regulators and voluntary frameworks do recommend an overview of your agents. What do you record for each agent, how does the register differ from the AI register and the record of processing, and how do you keep it current?
Read moreAn AI agent can take a wrong action, not just give a wrong answer, and hidden instructions in an email or web page will probably never be fully preventable. This article explains the lethal trifecta, OWASP's list for agents and the real cases, and what the GDPR, the AI Act and the Dutch Cybersecurity Act ask of you.
Read moreYou do not need a separate AI bureaucracy. The record of processing activities, the DPIA, privacy by design, the DPO, data processing agreements and the breach process together form the chassis for the duties under the AI Act. This is how to run AI governance on the privacy function you already have: six building blocks, one register, a RACI, the limits of the DPO's role, the Dutch public sector as a worked example and a 90-day starting plan.
Read moreWho is responsible for the AI Act in an organisation? Legally, the organisation in its role as provider or deployer. The evidence, however, is produced by IT, data and MLOps. The extended English version of Zahed Ashkara's AG Connect article, updated to the state of play on 8 October 2026.
Read moreA transparent benchmark of 61 EU AI Act questions across two AI search models, measuring how often Praxikon, Embed AI and LearnWize were cited as sources or named in answers.
Read moreAgentic AI governance means applying the EU AI Act and GDPR to AI agents that act autonomously. This hub brings the whole picture together: whether agents fall under the Act, your role in the value chain, the controls autonomous agents need, how to assess their risk, and who is responsible when an agent makes a mistake.
Read moreAI agents that execute tasks autonomously fall squarely under the EU AI Act through the definition in Article 3(1). The complete guide: the risk ladder applied to agents, the roles in the value chain, GDPR Article 22 as the boundary that applies today, and a governance framework covering permissions, oversight, logging and a kill switch.
Read moreThe AI Act, NIS2 and DORA largely regulate the same underlying capabilities: risk management, incidents, third parties, governance and logging. Here is how to build a single control set where each law becomes a view, instead of running three parallel compliance programmes.
Read moreThe EU AI Act does not mandate any specific compliance officer. Responsibility sits with the organisation itself, and therefore with the board. Here is a workable division of roles across the board, process owners, the DPO, the CISO and compliance.
Read moreAn AI agent that acts autonomously can make mistakes with real consequences. Who carries the responsibility? The chain explained: model provider, platform provider and deployer, plus what boards and legal counsel should assign internally today.
Read moreThe EU AI Act never literally requires an internal AI register, yet without an inventory you cannot demonstrably comply with almost any obligation. What the law actually demands, what regulators expect and how to build a register that stays alive.
Read moreDoes an ISO/IEC 42001 certificate make your organization EU AI Act compliant? No, but it is the best foundation you can build. What the standard actually certifies, where it overlaps with the AI Act and where the certificate stops.
Read moreDigital autonomy determines how organizations control AI decisions, manage vendor lock-in, and ensure human oversight remains meaningful in practice.
Read moreDutch DPA warns the proposed police surveillance law lacks clear limits, risking mass monitoring of innocent citizens without suspicion. Key findings.
Read moreThe Dutch DPA publishes RAN 6: 4 of 9 indicators are now red. AI in recruitment, transparency, and AI Act preparation fall short.
Read moreThe Pentagon demands that Anthropic remove two safety limits: no mass surveillance of Americans, no fully autonomous weapons. Dario Amodei refuses.
Read moreMore than 100 international experts, led by Turing Award winner Yoshua Bengio, published the most comprehensive AI safety report to date.
Read more80% of Fortune 500 firms use AI agents, but only 1 in 5 has mature governance. This 2026 guide covers what controls enterprises need to implement.
Read moreThe GPAI Code of Practice Signatory Taskforce sets the rules for AI models like GPT and Gemini. What does this mean for organizations building or using AI?
Read moreAnthropic research shows AI assistants can disempower users in edge cases. Learn what triggers it, which roles are most at risk, and how to prevent it.
Read moreOver 70% of banks use agentic AI, but governance lags behind. This 2026 guide covers what financial institutions must implement before supervisors act.
Read moreWhen Article 49 makes registration mandatory, who registers (provider or deployer), the Annex III scope, and how the not-high-risk claim under Article 6(3) works.
Read moreEU Commission's first draft Code of Practice on AI content transparency sets new labeling rules for AI-generated and AI-manipulated content.
Read moreAGI is not a well-defined concept, but a spectrum of increasingly broad and autonomous AI systems.
Read moreAI alignment sounds like a technical topic for labs and researchers, but in practice it affects executives, regulators and product teams daily: does an.
Read moreEU Commission's December 2025 consultation on AI regulatory sandboxes sets the framework. Here's what organizations need to prepare right now.
Read moreSeven lawsuits against OpenAI expose how ChatGPT allegedly encouraged vulnerable users toward suicide and reinforced delusional thinking.
Read moreThe European Data Protection Supervisor published a revised version of its GenAI guidance on October 28, 2025.
Read moreCEN and CENELEC have taken exceptional measures to deliver core standards for the EU AI Act faster.
Read moreUntil November 7, 2025, the European Commission is requesting feedback on the draft guidance and reporting template for reporting serious AI incidents.
Read moreThe EU AI Act is getting a Scientific Panel of 60 independent experts who will lay the technical foundation for policy and supervision from 2026.
Read moreDutch DPA tested four AI chatbots as voting guides and found more than half gave incorrect recommendations. What this means for AI in public discourse.
Read morePenalty provisions are active since August 2025, but most organizations aren't ready. Here's where enforcement readiness stands and what gaps remain.
Read more2025 marks a pivotal year in AI Governance: from experimental frameworks to operational compliance.
Read moreFrom a ‘voluntary’ GPAI Code to a practical vendor‑assurance process with clear evidence, workable obligations, and a realistic path to EN standards.
Read moreEU published a Code of Practice for general-purpose AI (GPAI). Model clauses that help deployers embed transparency, copyright, and safety in contracts.
Read moreEU AI Act mandates a public summary of training content. The Commission's template shows what providers must disclose—and which risks to avoid.
Read moreWhy the Commission withdrew the AILD in February 2025, and how the Product Liability Directive and AI Act now cover AI damage claims in the EU.
Read moreThe European Commission has officially recognized the voluntary code of practice for AI models as a legitimate compliance instrument under the AI...
Read moreThe algorithmic confidentiality privilege (ACP) is legal adviser Zahed Ashkara's proposal to give conversations with AI chatbots the same legal confidentiality as conversations with a doctor or lawyer. Published as an opinion piece in Het Financieele Dagblad, fully developed here.
Read moreOn July 10, 2025, the European Commission published the definitive General-Purpose AI Code of Practice.
Read moreEU Commission published the definitive Code of Practice for general-purpose AI. New obligations on transparency, safety, and copyright are now in effect.
Read moreWhy can we often not understand AI? This blog dives into the 'black box' of AI and explains why explainability (XAI) is essential for trust, fairness.
Read moreEU AI Act sandboxes give organizations a supervised testing environment for high-risk AI. Learn who qualifies, how to apply, and what benefits await.
Read moreHow do we as humans maintain grip on AI? This blog discusses concrete ways to maintain control, with examples and tips based on the EU AI Act.
Read moreHow do we stay in control of smart technology? This blog explores practical methods to maintain human control in AI use, with concrete examples and...
Read moreA comparative analysis of how we underestimate AI, similar to how we once underestimated electricity, and why we must take action now to embrace this...
Read moreWhen an AI system makes a legal error, who is liable? From algorithmic biases to privacy dilemmas - we dive into the ethical gray areas of AI in legal...
Read moreFrom boardroom to workplace: practical guidelines for implementing responsible AI systems within your organization.
Read moreAn analysis of how the EU AI Act will influence different sectors and businesses.
Read more