Dutch authority publishes AI literacy implementation guide

The Autoriteit Persoonsgegevens revealed a complete information on October 23, 2025, outlining sensible steps for organizations to construct enough AI literacy amongst workers and contractors working with synthetic intelligence techniques. The second information follows preliminary suggestions revealed in January 2025 and comes as enforcement of the EU AI Act approaches vital milestones.

In line with the AP, organisations that develop or deploy AI should guarantee enough data, expertise and understanding amongst people working these techniques. This obligation applies to each workers and exterior events deploying AI techniques on behalf of an organisation, together with contractors, service suppliers and purchasers. The requirement entered into power as a part of the broader AI Act framework, with enforcement obligations for general-purpose AI fashions starting August 2, 2025.

The information emerged from analysis performed between April and June 2025. The AP launched a name for enter in April, in search of responses from Dutch organisations about their approaches to AI literacy. Outcomes from this session revealed that many organisations stay in early levels of creating complete AI literacy programmes. Survey information confirmed 54% of responding organisations reported AI literacy on their administrative agenda, whereas 35% indicated partial consideration to the subject and 11% reported no present focus.

The European Fee offered clarification on AI literacy necessities by means of questions and solutions documentation launched in mid-2025. In line with this steerage, no single set of measures ensures ample AI literacy throughout all contexts. The suitable strategy depends upon three elements: the technical data and expertise of these working with AI techniques, the surroundings the place techniques function, and the people or teams affected by these techniques.

The AP’s framework establishes 4 distinct phases for constructing AI literacy. The identification part requires organisations to find out which AI techniques function inside their construction and assess related dangers and alternatives. This consists of documenting the diploma of autonomy every system possesses, with greater autonomy ranges requiring elevated literacy focus. Organisations should map present data ranges amongst workers working with totally different techniques and establish alternatives for AI deployment inside their operations.

Documentation necessities signify a vital element of demonstrating compliance. In line with the information, organisations should keep inside data monitoring their strategy and progress towards enough AI literacy. This documentation might show important throughout enforcement actions, as inadequate coaching or steerage may contribute to incidents involving AI techniques. The framework doesn’t require measuring literacy ranges for particular person workers or acquiring formal AI literacy certifications.

The goal-setting part emphasises prioritising concrete, measurable targets based mostly on danger ranges. Excessive-risk AI techniques require specific consideration, with the AI Act mandating that deployers guarantee staff with applicable experience oversee such techniques. The information notes that totally different organisational roles require distinct data and expertise past present technical capabilities. A developer would possibly possess sturdy technical skills however restricted consciousness of moral implications, highlighting the necessity for multidisciplinary data.

Organisations ought to set up centres of experience or appoint devoted specialists whom workers can seek the advice of relating to AI system questions. The AP emphasises creating secure, open work cultures the place workers really feel inspired to be taught from one another’s questions and errors. Survey information revealed that AI literacy usually develops by means of bottom-up initiatives, with particular person workers recognising significance and bringing consideration to administration. Nonetheless, the AP stresses that sustainable embedding requires top-down administration steerage and board-level involvement.

The implementation part addresses sensible execution of AI literacy methods. Board members should perceive AI literacy significance for correct embedding and budgeting. As a rule, AI literacy approaches ought to combine into broader AI technique or imaginative and prescient paperwork relatively than standing alone. Coaching and consciousness programmes kind the muse, with awareness-raising about deployment and improvement penalties offering a base for growing worker data and expertise.

Transparency inside organisations and externally about AI system utilization permits workers and concerned events to undertake vital and proactive stances. Monitoring mechanisms guarantee organisations can reveal AI literacy by means of documented progress. Survey outcomes indicated that AI literacy appeared on administrative agendas for almost all of responding organisations, although implementation diverse considerably.

The analysis part recognises that AI literacy represents an ongoing course of relatively than a hard and fast endpoint. Structural analysis requires monitoring and common evaluation of whether or not targets are being met, with outcomes informing changes to present measures. AI literacy ought to turn into a part of administration reporting buildings. As AI utilization inside organisations grows, corresponding will increase in AI literacy maturity turn into essential.

Survey information from the April name for enter revealed vital gaps in monitoring practices. Outcomes confirmed 33% of organisations reported no monitoring of AI literacy improvement, 45% indicated partial monitoring, and 22% confirmed energetic monitoring techniques. Measuring impression and data ranges stays below improvement, with organisations presently experimenting with baseline measurements and analysis codecs.

The information emphasises that AI literacy obligations prolong past high-risk techniques to all AI deployments. The AP serves as coordinating supervisor for algorithms and AI whereas contributing to preparation of AI Act supervision. Coverage, greatest practices and pointers on AI literacy proceed improvement at European degree, with the AP contributing to those efforts. A key requirement for accountable algorithm use includes understanding alternatives and impacts of such techniques, no matter whether or not they fall below AI Act high-risk classifications.

The AP performed intensive stakeholder engagement earlier than publishing this second information. Following the April name for enter, the authority organised a seminar titled “Working collectively on AI literacy” in June 2025. Insights from each initiatives knowledgeable the sensible examples and clarifications included within the multiannual motion plan outlined within the information.

Trying forward, the AP designated AI literacy as a spotlight space for ongoing supervision. The authority plans to discover AI literacy implementation standing amongst Dutch organisations all through coming months. In 2026, the AP will conduct a focused survey amongst Dutch organisations to watch AI literacy progress. Moreover, the authority introduced plans to organise a second seminar on AI literacy, offering platforms for organisations to trade data and experiences.

The publication coincides with broader European regulatory developments affecting AI deployment. The European Fee launched complete pointers on July 18, 2025, clarifying obligations for suppliers of general-purpose AI fashions. These pointers addressed mannequin classification standards, supplier identification, open-source exemptions and enforcement procedures, with compliance necessities coming into software on August 2, 2025.

Denmark positioned itself as a regulatory leader by turning into the primary EU member state to undertake nationwide laws implementing AI Act provisions. On Could 8, 2025, the Danish Parliament handed complete laws establishing governance frameworks required for AI regulation enforcement. The Company for Digital Authorities, Danish Knowledge Safety Authority and Danish Courtroom Administration function designated competent authorities below this framework.

The necessity for systematic AI literacy has gained recognition throughout expertise sectors. Zapier announced on May 30, 2025, that each one new workers should reveal AI fluency earlier than becoming a member of the organisation. This coverage represents one of the crucial complete AI literacy necessities carried out by a significant expertise agency, positioning AI fluency as elementary competency equal to literacy or numeracy.

Analysis on AI adoption patterns reveals demographic variations affecting literacy wants. A German study published in August 2025 confirmed generative AI adoption varies sharply by age, gender and academic background. These findings recommend organisations should develop diverse coaching approaches that meet staff at totally different technical familiarity ranges relatively than assuming uniform literacy.

The AP’s strategy displays rising recognition that AI deployment with out ample literacy creates compliance and operational dangers. With generative AI utilization growing in schooling and providers sectors, organisations specific considerations about workers utilizing AI purposes with out enough oversight, frameworks or safeguards. The information addresses this “concern of shedding management” by offering structured approaches to constructing organisational capability.

The Dutch authority’s concentrate on AI literacy aligns with broader European priorities of creating AI governance requirements whereas sustaining technological competitiveness. Organisations offering or deploying AI techniques should guarantee people have applicable expertise, data and understanding to deploy techniques responsibly. This allows organisations to maximise advantages from AI techniques whereas preserving elementary rights and values.

The multiannual motion plan outlined within the information offers prerequisite foundations together with managerial dedication, devoted finances allocation, clear organisational possession and accountability, and periodic reporting and monitoring processes. These stipulations help the iterative four-phase strategy throughout identification, aim setting, implementation and analysis levels.

Survey outcomes highlighted challenges organisations face in sustaining AI literacy. Continuity and safeguarding signify necessary however difficult duties, with solely a small portion of organisations actively monitoring improvement and progress. Measuring impression and data ranges stays below improvement, with organisations experimenting with baseline measurements and analysis codecs to trace effectiveness.

The AP notes that differentiation between capabilities, danger profiles and expertise ranges helps make clear data expectations for particular person workers and organisations general. Organisations emphasise the necessity to combine AI literacy into outlined roles, decision-making processes and present programmes corresponding to cybersecurity consciousness coaching or compliance frameworks.

The growing use of AI purposes corresponding to chatbots mechanically producing responses to scholar or shopper questions drives the urgency for correct governance and management. The AP underlines that organisations should take accountability for fostering and sustaining AI literacy, with developments within the subject evolving quickly whereas approaches to literacy creating much less shortly. The authority stresses that growing AI use requires corresponding will increase in AI literacy dedication.

Timeline

5 Ws Abstract

Who: The Autoriteit Persoonsgegevens, the Dutch information safety authority, revealed steerage affecting organisations offering or deploying AI techniques, their workers, contractors, service suppliers and purchasers working with AI techniques throughout the Netherlands and European Union.

What: A complete second information titled “Constructing AI literacy” establishing a four-phase implementation framework comprising identification, aim setting, implementation and analysis levels, with sensible examples clarifying the multiannual motion plan for guaranteeing enough AI literacy amongst people working or deploying AI techniques.

When: Revealed October 23, 2025, following a name for enter performed in April 2025 and a stakeholder seminar held in June 2025, with obligations below the AI Act coming into software on August 2, 2025, and deliberate follow-up survey scheduled for 2026.

The place: The Netherlands, with implications for organisations working all through the European Union below the AI Act framework, coordinated by the AP because the coordinating supervisor of algorithms and synthetic intelligence.

Why: Organisations should guarantee AI literacy amongst workers as a authorized obligation below the AI Act whereas addressing the fact that approaches to AI literacy are creating much less quickly than AI expertise itself, with survey information exhibiting many organisations stay in early implementation levels regardless of rising AI utilization in sectors like schooling and providers.