Since 2 February 2025: AI literacy (Article 4)
Article 4 of the Act has applied since 2 February 2025. It obliges providers and deployers of AI systems to take measures to ensure, to their best extent, a sufficient level of AI literacy among their staff. The provision takes into account the technical knowledge, experience and training of the people involved, as well as the context in which the systems are used.
The Act does not prescribe a specific format. It describes an outcome: employees who use AI tools should be able to judge what those tools can do, where their limits are, and what risks their use carries. Whether existing measures are sufficient is a question of the individual case.
From 2 August 2026: transparency duties (Article 50)
Article 50 governs transparency towards the people who come into contact with AI. From 2 August 2026, among other things: anyone deploying an AI system that interacts with people must inform them that they are dealing with a machine, unless that is obvious from the circumstances.
AI-generated or AI-manipulated image, audio and video content that deceptively resembles real persons or events must be labelled as artificially generated. The provision also foresees labelling duties for AI-generated text published to inform the public.
From 2 December 2027: the high-risk regime (Annex III)
The Act classifies as high-risk, among others, AI systems used in the employment context: in recruitment and applicant screening, in decisions on promotion and termination, in task allocation, and in monitoring and evaluating employees. Such systems carry extensive obligations, from risk assessment through human oversight to documentation.
The start of application of this regime was deferred to 2 December 2027. Companies that use such systems, or plan to introduce them, have more time as a result, but not a smaller set of obligations.
Independent of Brussels: the works council (§ 87(1) no. 6 BetrVG)
Alongside the European regulation stands German works constitution law. Under § 87(1) no. 6 of the Works Constitution Act (BetrVG), the works council has a co-determination right when technical systems are introduced or used that are intended to monitor the behaviour or performance of employees.
Under the settled case law of the Federal Labour Court, the employer’s intent is not decisive. It is sufficient that the system is objectively suitable for monitoring. Many AI tools log inputs and usage and can therefore fall under the provision. Introducing such a system without involving the works council can be invalid and can trigger injunctive claims.
The common denominator
All four points rest on the same foundation: a company has to know which AI is used in the organisation, for what, and under which rules. That is exactly where our approach begins.