Unacceptable risk Banned
Practices such as social scoring or behavioural manipulation. Prohibited in the EU since February 2025.
EU AI Act
Your team already uses AI. The EU now requires it to know how.
Since 2 February 2025, the EU AI Act requires organisations that use AI to ensure a sufficient level of AI literacy among their people. We prepare the team working with ChatGPT, Claude, Copilot or Perplexity to use AI in a compliant, safe way, without slowing down.
Book a discovery call
30-60 min. We see if AI is the right tool for your organisation.
What the Act requires
A simple obligation, with real consequences
Article 4 requires providers and organisations that use AI to make sure their people understand the capabilities, limits and risks of the tools they use. The Act sorts AI systems into four risk levels, and prohibited practices carry fines of up to €35 million or 7% of global annual turnover.
High risk
AI in recruitment, credit, education, health or infrastructure. Requires documentation, human oversight and a conformity assessment.
Limited risk
Chatbots and generated content. Transparency duty: people must know they are interacting with AI.
Minimal risk
Most everyday uses. No specific obligations, but AI literacy is still required for the people using the tools.
The compliance training
Two tiers, matched to how far you are with AI
01 / Compliance essentials
Tier 1 · Compliance essentials
The session that brings the team to the minimum Article 4 requires: what the Act is, what falls on you as an organisation that uses AI, and how to work day to day without creating risk.
What it covers
- What the EU AI Act is and what it means for you in practice
- The four risk levels and which practices are banned
- The capabilities, limits and risks of the tools you use (ChatGPT, Claude, Copilot, Perplexity)
- Confidentiality, personal data and the overlap with GDPR
- How to recognise and verify an AI output before you use it
- The transparency duty: when you must disclose that content is AI-generated
What the team gets
- Certificate of completion for every team member
- AI training record for the organisation - the evidence Article 4 expects
- A one-page internal guide to safe AI use
02 / Compliance + practice
Tier 2 · Compliance + practice
Everything in Tier 1, plus exercises on your own tools and use cases, a prompting and verification workshop, and the first steps toward an internal AI usage policy.
What it covers
- Everything in Tier 1
- Classifying your own use cases by risk level, together with the team
- Prompting workshop: how to write better, safer prompts
- Verifying outputs: techniques to control hallucinations and errors
- Clear rules for confidential and personal data in AI tools
- An internal AI usage policy skeleton, started in the session
What the team gets
- Certificate of completion for every member
- AI training record for the organisation
- Internal AI usage policy template
- A risk-level map of your use cases
- A prompting and verification checklist
The Act does not prescribe a fixed curriculum: training has to fit your roles, tools and context. Both tiers are tailored this way.
Manager self-assessment
Does your team need training now?
Six questions, under two minutes. You get a clear priority and concrete next steps. Nothing leaves your device - the scoring runs locally in your browser.
Question 1 of 6
How much does your team already use AI tools (ChatGPT, Claude, Copilot, Perplexity)?
Are AI outputs checked by a person before they are used or sent on?
Do you have a written internal policy for using AI?
Do you know which EU AI Act risk category your use of AI falls into?
Has the team had any training on responsible AI use or the EU AI Act?
In what context do you use AI most?
High priority
Schedule training soon
Your team uses AI in ways that clearly fall under the Act, but without the review, policy or training to protect you. The AI literacy duty (Article 4) has been in force since 2 February 2025.
Concrete next steps
- Schedule a 2–4 hour compliance session for the whole team in the next 4–6 weeks.
- Make an inventory of the AI tools in use and the kind of data that goes into them.
- Classify each use case against the Act's four risk levels.
- Set a firm rule: no AI output reaches a client or a decision without human review.
- Start a written training record - it is the main evidence of compliance Article 4 asks for.
Medium priority
Training recommended within 6 months
You already have a few good practices, but with gaps that can expose you. Supervision and enforcement begin on 2 August 2026, so you have time to close them without pressure.
Concrete next steps
- Plan a compliance training before August 2026.
- Turn the informal rules into a written, one-page internal policy.
- Decide clearly who reviews AI outputs in sensitive tasks (personal data, legal documents, external communication).
- Run a short prompting and verification session for the roles that use AI daily.
- Document what training each role has had, so you can demonstrate compliance.
Low priority
You're in a good position
Your team uses AI with rules, review and a good level of risk awareness. The goal now is to keep that level as the tools and the legal framework evolve.
Concrete next steps
- Keep the training record up to date and rerun a short session whenever a major tool or law changes.
- Review the internal AI usage policy once a year and after each new tool you adopt.
- Name a person responsible for AI compliance who tracks the Act's deadlines.
- Run a prompting and verification refresher for new joiners.
- Reassess the risk level if you extend AI into new areas (HR, credit, health, public services).
Book a discovery call
30-60 min. We see if AI is the right tool for your organisation.
Who delivers it
Compliance taught from practice, not theory
The founder, Ștefan Rusu, who holds a PhD in finance and AI (Summa Cum Laude), worked in the European Central Bank's Digital and AI Transformation programme and trained more than 400 experts in a strictly regulated environment. Compliance, confidentiality and responsible AI use are part of every programme we deliver.
The training programmes were delivered during the founder’s time working at the European Central Bank. This does not necessarily constitute a public endorsement of AI Solutions or its services by the European Central Bank.
Frequently asked questions
Before you write to us
- Who does the obligation apply to?
- Both providers of AI systems and organisations that merely use them (deployers). If your team uses AI tools in its daily work, you fall into this category.
- What does a “sufficient” level of AI literacy mean?
- The Act does not set a fixed curriculum or an exam. Training has to fit the roles, experience and context in which you use AI. That is why both tiers are tailored to your team.
- Do we get certificates and documents?
- Yes. Each participant receives a certificate, and the organisation receives a training record that evidences the instruction, exactly what Article 4 asks for. Tier 2 adds an internal policy template and a risk-level map of your use cases.
- Do we need training even if we use AI only occasionally?
- Yes, but at a lower intensity. Tier 1, at 2–4 hours, is enough for most teams with light use.
- Is this legal advice?
- No. It is practical compliance training, delivered from experience in regulated environments. For specific legal interpretation, we work alongside your legal counsel.
Let's see what training your team needs
A 30 to 60 minute conversation. We tell you which tier fits your team and what documents you receive, with no obligation.
Book a discovery call
30-60 min. We see if AI is the right tool for your organisation.