European Parliament blocks AI on lawmakers’ devices, citing security risks

The European Parliament has taken a decisive step to prevent lawmakers from utilizing built-in AI tools on their work devices, citing significant cybersecurity and privacy concerns. An internal email shared with Politico revealed that the parliament's IT department expressed doubts about the security of data uploaded to the servers of AI companies, stating that the risks associated with sharing confidential correspondence are still under evaluation. The email outlined that, due to these uncertainties, it is deemed safer to disable AI functionalities for now. This decision relates particularly to popular AI chatbots such as Anthropic’s Claude, Microsoft’s Copilot, and OpenAI’s ChatGPT, which can expose users to potential data breaches. If sensitive information is uploaded, U.S. authorities could compel these companies to hand over user data, increasing the risk of unintended information sharing among users. Europe is recognized for its stringent data protection regulations. However, last year, the European Commission proposed legislation that could ease these protections, allowing tech giants greater access to European data for training AI models. Critics argue that such measures compromise user privacy and cater to U.S. tech interests. This restriction on AI usage comes at a time when various EU member states are reassessing their ties with U.S. technology firms, which must adhere to U.S. law and its fluctuating political environment. Recent actions by the U.S. Department of Homeland Security, including numerous subpoenas for information from tech companies regarding critics of the Trump administration, have raised alarms. Firms like Google, Meta, and Reddit have complied with these requests, despite the absence of judicial oversight, further emphasizing the precarious nature of data security in the current landscape.