Metsola tells MEPs not to rely on AI when drafting laws

European Parliament President Roberta Metsola has cautioned Members of the European Parliament (MEPs) against relying excessively on artificial intelligence (AI) when drafting laws. Speaking at an event on the use of AI in the Parliament, Metsola emphasized that while AI can assist in legal research, it cannot replace the critical judgment and debate necessary in lawmaking.

She stressed that MEPs must maintain "democratic accountability" and ensure transparency in the use of AI. The Parliament is developing its own AI tool for sensitive work hosted on internal infrastructure, but commercial AI models like ChatGPT, Claude, and Meta's Llama are still permissible for less sensitive tasks. The administration plans to release new guidelines on AI usage soon, building upon existing 2024 guidelines.

Chiocchetti highlighted an incident where an MEP incorrectly relied on AI for legal advice, which was disproven by Parliament's jurists. The new guidelines stress openness about AI use, avoidance of sensitive material in public tools, and thorough checks before finalizing any document. Metsola warned that delegating political debate to AI risks undermining the responsibility of elected officials.

She stressed that each piece of legislation should be authored, vetted, and defensible by those who voted on it, ensuring accountability in the legislative process.