Europe recalibrates AI rules
Coverage argues Europe is stepping back from maximalist AI regulation while keeping the AI Act framework in place, and legal analysis notes the EU withdrew the separate AI Liability Directive, narrowing the liability superstructure. The result described is a more selective regulatory posture that preserves strict rules for high‑risk systems while easing some parallel liability ambitions. (thenextweb.com, mondaq.com)
Europe is not scrapping its artificial intelligence rulebook. It is keeping the European Union Artificial Intelligence Act while dropping a separate liability bill that would have made lawsuits over artificial intelligence easier. (digital-strategy.ec.europa.eu, europarl.europa.eu) The core law is already on the books: Regulation (EU) 2024/1689 entered into force in August 2024, and most of it applies from August 2, 2026, with some bans and governance rules already phased in earlier. The European Commission says the law uses a risk-based system, with outright bans for a narrow set of uses and stricter duties for high-risk systems. (digital-strategy.ec.europa.eu, digital-strategy.ec.europa.eu, artificialintelligenceact.eu) The measure that fell away was the Artificial Intelligence Liability Directive, a September 28, 2022 proposal that would have added a presumption of causality and broader evidence disclosure in some civil cases. The European Parliament’s legislative tracker says work on that file was tied to the Artificial Intelligence Act and later resumed after the main law was adopted. (eur-lex.europa.eu, europarl.europa.eu) The European Commission’s 2025 work programme, published on February 11, 2025, put simplification and competitiveness at the center of the new mandate. The programme listed 37 proposals for withdrawal, and the Official Journal later recorded the withdrawal of pending proposals approved at the Commission’s July 16, 2025 meeting. (commission.europa.eu, eur-lex.europa.eu) That leaves Europe with a narrower legal stack than many companies expected in 2022. The compliance burden now sits more heavily inside the Artificial Intelligence Act itself and in existing product safety and liability rules, rather than in a second artificial intelligence-specific civil liability regime. (digital-strategy.ec.europa.eu, eur-lex.europa.eu) The shift also lines up with a broader change in Brussels from rule-writing toward industrial policy. A September 2025 briefing from the European Parliament Research Service said the Commission’s April 2025 “AI Continent Action Plan” looked beyond regulation to investment, infrastructure, skills, and adoption. (europarl.europa.eu) The debate inside Europe has not gone away. The same Parliament briefing said industry groups want simpler rules to speed deployment, while civil society groups warn against weakening safeguards on safety, fundamental rights, and accountability. (europarl.europa.eu, digital-strategy.ec.europa.eu) So the recalibration is not a repeal. Europe is still enforcing a detailed artificial intelligence law, but it is no longer building the fuller liability superstructure that once looked set to sit beside it. (digital-strategy.ec.europa.eu, eur-lex.europa.eu)
