EU AI Act training that makes sense at work.
The EU AI Act places AI-literacy responsibilities on organisations that provide or use AI systems. Adapt Practice helps teams understand what that means — and turn it into safer, more confident everyday AI use.
Practical workplace training for employees, managers and non-technical teams.
If your people use AI at work, AI literacy is now an organisational responsibility.
Many organisations are already using AI systems — often without having made a single formal decision to "adopt AI".
It arrives through the tools staff already have: a chat assistant opened in a browser, a Copilot button in office software, or an AI feature quietly added to an existing platform. Under the EU AI Act, an organisation using AI systems like these in its work will commonly be acting as a deployer — and deployers have AI-literacy responsibilities under Article 4.
Exactly which obligations apply still depends on the systems involved and how they are used. Not every tool creates the same obligations, and higher-risk uses bring additional requirements.
Where AI often shows up at work
- ChatGPT, Gemini and similar AI assistants
- Microsoft Copilot and AI-assisted office software
- Recruitment and HR screening tools
- Customer-service chatbots and support systems
- Automated decision-support and scoring tools
- AI features embedded in existing SaaS products
Provider
Builds or supplies the AI system
An organisation that develops an AI system — or has one developed — and places it on the market or puts it into service under its own name or trademark.
Example: a software company releasing an AI screening product.
Deployer
Uses an AI system in its work
An organisation using an AI system under its authority, other than for purely personal, non-professional activity.
Example: a firm whose staff use Copilot to draft reports, or whose HR team uses an AI screening product.
For many ordinary organisations, the relevant role will be deployer rather than provider. Some organisations may be both.
What does Article 4 actually require?
In plain English: if your organisation acts as a provider or deployer of AI systems, it must take measures to support the AI literacy of the staff and others who operate or use those systems on its behalf.
Those measures should reflect who the people are, what they already know, the context in which the systems are used, and who could be affected. The law does not require every individual to reach a guaranteed level of AI literacy.
“Providers and deployers of AI systems shall take measures to support the development of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used. This obligation does not require providers or deployers to guarantee any specific level of AI literacy of any individual.”
Four practical steps, based on European Commission guidance
-
Understand
Build a general understanding of AI: what it is, broadly how it works, which systems the organisation uses, and their opportunities and limitations.
-
Identify
Establish the organisation's relationship with AI — provider, deployer, or potentially both — and which systems that applies to.
-
Recognise risk
Help people understand the risks relevant to the systems and tasks they actually deal with.
- Unreliable output and hallucination
- Over-reliance and automation bias
- Personal and confidential information
- Bias and discrimination
- Transparency and human oversight
- Cyber and social-engineering risks, where relevant
-
Adapt to the role
Shape training and guidance around the people receiving it.
- Job role and prior knowledge
- The AI tools actually used
- Decisions being made with AI
- People who could be affected
- The level of risk involved
There is no statutory 60-minute course.
The AI Act does not prescribe a fixed duration, a universal syllabus, a compulsory examination or a mandatory Article 4 certificate. The European Commission's guidance is clear that there is no one-size-fits-all approach: the appropriate measures depend on the organisation and how it uses AI.
Training is one practical way to build that capability — and to keep a record of the action taken.
Practical AI literacy — not a legal lecture.
Employees do not need somebody reading regulations at them.
Sessions are interactive and connect the legislation to realistic situations from your organisation's working day. Participants leave understanding:
- Explain — a short, plain-English explanation.
- Demonstrate — a live example on a realistic workplace task.
- Practise — a short scenario, done straight away.
- Apply — relate it to people's own roles and rules.
- A plain-English EU AI Act overview
- Your organisation's role and each employee's part in it
- Everyday workplace AI risks
- Practical scenarios
- Data and information handling
- Hallucinations and verification
- Automation bias and human oversight
- Responsible AI behaviour
- Your organisational rules and policies
- Role-relevant discussion
- Questions and action points
- Focused staff briefing
- Interactive team workshop
- Manager or leadership session
- Role-specific training
- Tailored organisational programme
Format and duration are tailored to the organisation, audience and AI use.
A practical baseline for workplace AI literacy
This is Adapt Practice's practical interpretation of a sensible baseline for everyday workplace AI users, drawing on the legislation and European Commission guidance. It is not a legally mandated syllabus — the right content depends on your people, systems and risks.
-
AI fundamentals
What AI is, how generative AI differs from traditional automation, and why AI output is a probabilistic prediction rather than guaranteed truth.
-
Capabilities and limitations
What these tools do well, where they hallucinate or miss context, and why a confident answer is not the same as a correct one.
-
Data and confidentiality
What should and should not be entered — personal data, confidential and commercially sensitive information — and why organisation-approved tools matter.
-
Human judgement
Checking evidence, recognising over-reliance, knowing when human review is essential, and when to stop and escalate.
-
Bias, fairness and affected people
Why outputs can be biased, the added risk when AI affects people, and why context and oversight matter.
-
Responsible workplace use
Approved uses, transparency where appropriate, following organisational rules and knowing how to report a concern.
-
EU AI Act awareness
Provider and deployer roles, Article 4, the Act's risk-based approach, and why different roles need different depths of knowledge.
Different roles. Different risks. Different depth.
Article 4 expects measures to reflect people's knowledge, experience and the context in which AI is used. In practice, that usually means layers of training rather than one session for everyone.
-
Depth: foundation
Everyday AI users
People using tools such as Copilot, ChatGPT or AI-enabled workplace applications.
Focus
Safe everyday use, information handling, verification, limitations and organisational rules.
-
Depth: foundation + oversight
Managers & team leaders
People approving AI use, supervising staff or making decisions based on AI-assisted work.
Additional focus
Human oversight, accountability, escalation, affected people and appropriate use.
-
Depth: foundation + oversight + governance
Specialist & higher-risk roles
HR, recruitment, compliance, risk, IT, procurement, or teams working with AI that could materially affect people.
Additional focus
Risk classification, governance, documentation, procurement questions and role-specific controls.
Certain AI systems used in employment — for example for recruitment and selection, or for decisions on promotion, termination, task allocation or performance monitoring — are listed as high-risk under the AI Act and bring obligations beyond Article 4, including a requirement for deployers to assign human oversight to people with the necessary competence, training and authority. Following Regulation (EU) 2026/1744, those high-risk rules apply from 2 December 2027. This training can support preparation for that work, but it is not a complete high-risk AI compliance programme.
More than a slide deck.
Depending on the format agreed, an engagement can include the items listed here.
Records can support the organisation in documenting the AI-literacy measures it has taken. The European Commission notes that organisations can keep internal records of training and other guiding initiatives — no certificate is required.
This training provides practical education and compliance support. It is not legal advice, and the measures appropriate to an organisation will depend on its AI systems, roles, uses and risk context.
- Live, facilitated training
- Examples set in your organisation's context
- Participant materials and a reference guide
- An attendance or completion record
- Learning objectives and a training outline
- Practical responsible-use guidance
- Recommended next actions
- Optional role-specific follow-up
Article 4 training is a strong starting point — not the whole journey.
-
Understand
AI literacy and Article 4 awareness across the people using AI.
-
Equip
Practical, safe-use behaviours applied to everyday work.
-
Embed
Policies, governance, manager support and adoption practices.
Depending on the organisation and its AI use, responsible adoption may also involve:
- AI usage guidance or policy
- Manager briefings
- Governance arrangements
- An inventory of AI systems in use
- Risk assessment
- Procurement questions
- Data and security controls
- Role-specific processes
Not every organisation needs every item. For hands-on productivity skills, see Practical Workplace AI, or explore all services.
For organisations that want a sensible, practical response.
For
- Organisations beginning to use workplace AI
- Employers whose staff already use ChatGPT, Copilot or Gemini
- HR and L&D teams building AI capability
- Managers responsible for responsible AI use
- Organisations developing internal AI guidance
- Businesses seeking a practical response to Article 4
- Organisations needing different learning levels for different roles
This Is Not
- Legal advice
- An AI engineering programme
- An "EU-approved" certification — Article 4 does not require a certificate
- A generic compliance slideshow
- A substitute for wider governance where higher-risk systems are involved
How Article 4 got here.
-
The EU AI Act — Regulation (EU) 2024/1689 — entered into force.
-
Article 4 AI-literacy provisions began to apply.
-
Amended Article 4 wording took effect under Regulation (EU) 2026/1744, replacing the original requirement to ensure a "sufficient level" of AI literacy.
-
According to the European Commission, supervision and enforcement of Article 4 by national market-surveillance authorities applies from this date.
Official sources
- Regulation (EU) 2024/1689 — the EU AI Act (EUR-Lex) (opens in a new tab)
- Regulation (EU) 2026/1744 — Digital Omnibus on AI (EUR-Lex) (opens in a new tab)
- European Commission — AI talent, skills and literacy (opens in a new tab)
- European Commission — AI Literacy Questions & Answers (opens in a new tab)
- European Commission — Repository of AI literacy practices (opens in a new tab)
Common questions about EU AI Act training
Does the EU AI Act create an AI-literacy responsibility for employers?
Yes. Article 4 requires providers and deployers of AI systems to take measures to support the AI literacy of staff and others using or operating AI systems on their behalf. Training is one practical way to support that responsibility, although the Act does not prescribe one universal course or format.
Can Article 4 apply when employees use tools such as ChatGPT or Microsoft Copilot?
Yes. An organisation using AI systems supplied by others in its work will commonly be acting as a deployer. The European Commission's guidance specifically recognises that staff using tools such as ChatGPT for work should understand relevant risks, including unreliable or hallucinated output. The precise position still depends on the systems involved and how they are used.
Can AI literacy training be adapted for different roles?
Yes. Article 4 takes a contextual approach, considering people's technical knowledge, experience, education and training, the context in which AI systems are used, and the people who may be affected. Everyday users, managers and specialist roles may therefore need different levels of depth.
Can organisations choose a training format and duration that fits their needs?
Yes. The AI Act does not prescribe a fixed number of training hours. The appropriate approach depends on the organisation's people, AI systems, use cases and risk context.
Can organisations document their AI-literacy activity without an EU AI Act certificate?
Yes. Article 4 does not require a specific certificate. The European Commission says organisations can keep internal records of training and other measures taken to support AI literacy.
Can AI literacy training form part of an organisation's wider EU AI Act response?
Yes. Training can support an organisation's Article 4 responsibilities and strengthen responsible workplace AI use. It is only one part of the wider AI Act, however, and additional obligations may apply depending on the organisation's role and the AI systems it provides or uses.
Is Article 4 already in effect?
Yes. Article 4 has applied since 2 February 2025. Its wording was amended by Regulation (EU) 2026/1744, which entered into force on 27 July 2026. According to the European Commission, supervision and enforcement of Article 4 by national market-surveillance authorities applies from 2 August 2026.
Can Adapt Practice help without replacing legal advice?
Yes. Adapt Practice provides practical training and adoption support that helps organisations build AI literacy and safer workplace practices. Where a formal legal opinion is needed on specific obligations, organisations should also seek advice from a qualified legal adviser.
Can the training be tailored to our organisation?
Yes. Content can reflect your roles, the AI tools you use, your current policies, realistic use cases and your risk context.
Turn AI literacy into practical workplace capability.
If your organisation is already using AI — or preparing to — we can help you build the understanding, judgement and shared practices your people need.