The European Parliament has reportedly restricted lawmakers from using built-in artificial intelligence tools on official work devices, citing concerns over cybersecurity and data privacy.
According to an internal communication reported by Politico, the Parliament’s IT department said it could not fully ensure the safety of information uploaded to external AI providers’ servers. The email noted that the scope of data potentially shared with AI companies is still under review, leading officials to conclude that disabling such features is the more secure option for now.
The concern centers on cloud-based AI assistants such as Anthropic’s Claude, Microsoft’s Copilot, and OpenAI’s ChatGPT. When users upload documents or correspondence to these services, the data may be stored or processed on servers outside the EU. In some cases, companies operating under U.S. jurisdiction can be compelled by American authorities to provide user data.
There are also concerns that AI systems often use submitted information to refine and train their models, raising the possibility that sensitive material shared by one user could inadvertently influence outputs seen by others.
The decision comes against the backdrop of Europe’s stringent data protection framework. While the EU is known for some of the world’s toughest privacy regulations, the European Commission has recently proposed adjustments aimed at easing certain data restrictions to support AI development. Critics argue such measures could weaken safeguards and disproportionately benefit major U.S. technology firms.
The restriction also reflects a broader reassessment among EU institutions and member states regarding reliance on American tech companies, which remain subject to U.S. legal authority and shifting political dynamics.
Reports have highlighted recent actions by the U.S. Department of Homeland Security, which issued subpoenas to several U.S.-based technology platforms seeking user information in politically sensitive matters. Companies including Google, Meta, and Reddit have reportedly complied in certain instances, even when the requests were not backed by court orders.
The European Parliament’s move underscores growing caution within EU institutions about the risks associated with integrating third-party AI services into official government workflows.
