Metsola Warns MEPs Against Overreliance on AI in Legislative Drafting
European Parliament sets new guidelines to ensure democratic accountability in AI-assisted lawmaking processes.
1Legislative Integrity Concerns
European Parliament President Roberta Metsola has cautioned lawmakers against relying excessively on artificial intelligence for drafting legislation. She emphasized that AI tools cannot replace the fundamental necessity of constructive human debate or critical thinking. Metsola argued that democratic accountability remains a core responsibility for elected officials, regardless of technological assistance in drafting.
2Internal AI Infrastructure
Secretary-General Alessandro Chiocchetti announced that the Parliament is currently developing a proprietary AI tool designed specifically for sensitive parliamentary work. This system will be hosted entirely on internal infrastructure to maintain data security. Meanwhile, commercial models like ChatGPT and Llama remain permitted only for less sensitive, non-confidential administrative tasks.
3Risks of Hallucinations
Parliament officials have observed a significant increase in amendments generated by AI, some of which reference non-existent laws. Chiocchetti highlighted a specific incident where an MEP received incorrect legal advice from a chatbot. Parliamentary jurists required two full days of research to verify and prove the AI-generated information was factually wrong.
4Forthcoming Regulatory Guidelines
The administration plans to submit updated AI usage guidelines to senior parliamentary leaders in the near future. These new rules will build upon existing staff protocols established in 2024. The framework focuses on transparency, data security, and the absolute requirement for human verification before any document is officially submitted or finalized.
5Core Principles for Usage
The proposed guidelines mandate three primary requirements for all staff and MEPs. Users must be transparent about AI involvement, avoid inputting sensitive material into public tools, and personally verify all outputs. Because the official's name is attached to the final text, they bear full responsibility for the accuracy of the content.
6Managing Legislative Volume
Chiocchetti expressed interest in utilizing AI to assist with the administrative burden of sorting through thousands of individual amendments. While the technology offers efficiency for organizational tasks, the Parliament maintains that the actual drafting and political consideration of laws must remain strictly in the hands of elected members to ensure legitimacy.
7Democratic Accountability Standards
Metsola stressed that citizens must be able to identify who decided on specific regulations to hold them accountable. The Parliament fears that delegating political debate to chatbots could erode the democratic process. Every text submitted for a vote must be written, thoroughly checked, and defended by a human representative of the public.
8Future of Parliamentary Work
As the European Parliament integrates new technologies, the focus remains on balancing efficiency with institutional responsibility. By creating internal tools and strict oversight, the body aims to mitigate the risks of misinformation. The ultimate goal is to ensure that AI serves as a supportive instrument rather than a replacement for human judgment.