European lawmakers are pressing ahead with tougher AI safeguards, advancing amendments that would require more transparency and independent oversight for high-risk systems. The proposal comes as officials prepare for a global summit expected to set the tone for international AI policy.

Under the draft changes, companies developing or deploying sensitive AI tools could face real-time disclosure obligations and independent audits. Supporters say the measures are needed to reduce the risk of harmful errors, hidden bias, and opaque decision-making in systems used at scale.

The push reflects growing concern in Europe that AI regulation must move faster than the technology itself. Lawmakers are trying to balance innovation with public safety, while rights groups have warned that weak rules could leave workers, consumers, and marginalized communities exposed to abuse or discrimination.

A vote on the amendments is expected next week, and the outcome could shape the bloc’s stance in broader talks on global AI governance. If approved, the measures would mark another step toward a more enforceable regulatory framework for advanced digital systems.

Ai safety rules

EU lawmakers seek binding AI safety rules before global summit