EU Parliament rolls out a secure AI hub for lawmakers

The European Parliament is preparing to launch EPGenAI Hub, an officially approved AI platform meant to give lawmakers and staff a safer alternative to public chatbots. The system will support legislative work using models from Meta, OpenAI, Anthropic, and Mistral, amid mounting concern over errors, hallucinations, and low-quality AI-generated material showing up in parliamentary documents.

According to the report, about 20% of parliament staff already use AI tools in their daily work — roughly 2,100 people. That growing use has helped push the issue into the open, with some legislative materials reportedly containing mistakes and AI hallucinations.

“Parliament has not been immune to AI slop and misleading AI-generated content.”

Benifei, who led the European Parliament’s work on the AI Act, said MEPs should set an example as the law’s transparency rules start to apply.

At the moment, MEPs are not required to disclose when AI tools are used to draft or amend legislative materials. Barry Andrews, an Irish Renew lawmaker, told the source there is clear evidence of AI use across parliament.

Recommended reading

Ramp opens its LLM router after 30% cost cuts

“That is their choice and I fully understand given our workload. But for me personally, these are core democratic functions which voters elected us to perform.”

There are already guidelines for parliamentary employees. Politico reported that staff must label materials altered or produced with AI as “AI-assisted,” and they are barred from entering sensitive information into AI tools.

MEPs and staff have already been shown demos of EPGenAI Hub. The platform runs on both the parliament’s own infrastructure and external hosting, enabling access to models including Meta’s Llama, OpenAI’s GPT-OSS, Mistral Small, OpenAI’s ChatGPT, and Anthropic’s Claude Sonnet.