European Parliament committees have backed a proposal to postpone the application of certain obligations under the EU Artificial Intelligence Act, particularly those affecting high-risk AI systems. MEPs support giving organisations more time to prepare, but favour a shorter delay than originally proposed by the European Commission, suggesting a new deadline of November 2026 instead of February 2027.
The move aims to balance innovation and competitiveness with the need to ensure robust safeguards, technical standards and implementation guidance are fully in place before stricter rules take effect. At the same time, lawmakers reaffirmed their commitment to maintaining strong protections, including bans on harmful AI practices such as systems generating non-consensual intimate content.