The EU AI Regulation means business from December 2026 onward. Following the publication of an amendment regulation on 24 July 2026 and the general applicability date of 2 August 2026, companies now face concrete compliance deadlines that demand immediate action. The German Bar Association (Deutsches Anwaltsinstitut) is responding to the tightened legal situation with a digital seminar on 4 September 2026.
Key facts at a glance
- 2 December 2026: New prohibitions under Article 5 of the AI Act take effect
- 2 December 2027: Stricter requirements for high-risk AI systems (e.g. in personnel management) become relevant
- August 2028: Further binding provisions for high-risk systems follow
- 97 percent of companies rate AI as relevant (KPMG study 2026)
The timeline becomes concrete
Implementation of the EU AI Regulation is staggered. New provisions on AI competence under Article 4 have been in force since 2 February 2025. Transparency requirements for certain AI systems have been effective since 2 August 2026. But the critical dates still lie ahead:
| Deadline | Regulation | Affected parties |
|---|---|---|
| 2 December 2026 | Prohibitions under Article 5 | All AI providers |
| 2 December 2027 | High-risk requirements (Art. 6 para. 2) | Personnel management, assessment systems |
| August 2028 | Further high-risk provisions | Systems with significant impact |
Governance gap threatens implementation
The paradox: while 98 percent of surveyed companies have an AI strategy, only 39 percent actively steer their AI activities through top management. This governance gap is already taking its toll. According to a report by Optro, 40 percent of companies have already experienced faulty AI results, and 27 percent reported data protection violations. Those working without clear accountability will miss the new compliance requirements.
Employment law and works councils in focus
The German Bar Association's seminar places emphasis on the employment law dimension. Led by Dr. Marc Becker and Dr. Stefan Müller, central topics such as AI definition, the regulatory framework, liability issues, and – particularly important – works council co-determination rights will be discussed. The reason: AI systems in personnel management fall under the high-risk category and will be subject to stricter requirements from December 2027 onward. Works councils must be able to participate in shaping these systems.
Switzerland also moving closer
For companies with international operations: Switzerland is preparing a consultation draft by the end of 2026. Swiss companies with EU business activities must already comply with the AI Act's transparency requirements – a preview of possible harmonization.
What this means for you
The new deadlines are not theoretical regulations but an implementation roadmap. Companies should now classify their AI systems (high-risk yes/no?), clarify their governance structures, and anchor responsibilities. Those not prepared by December 2026 risk not only fines but also reputational damage. The odds are good: with clear strategy and top management engagement, compliance becomes a competitive advantage, not a burden.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
