European Parliament Disables AI Features on Work Devices Over Security Risks

The European Parliament has reportedly disabled built-in AI features on its corporate devices due to cybersecurity and data protection concerns.

According to Politico, lawmakers and staff received an internal email informing them about the precautionary move. Officials raised concerns that certain AI assistants rely heavily on cloud-based processing, even for tasks that could run locally, potentially exposing sensitive institutional data.

AI Features Disabled as Precautionary Measure

The Parliament confirmed that it is still assessing the full scope of data shared with external AI service providers. Until it fully understands how these tools handle internal documents and communications, leadership considers disabling AI features the safest course of action.

Staff members were also advised to exercise similar caution on personal devices used for work-related activities. The internal guidance specifically warned against granting AI tools permission to scan work documents, emails, or private communications. Lawmakers were further urged to carefully review third-party AI applications before installing them or allowing access to institutional data.

The European Parliament declined to specify which AI features, software platforms, or operating systems were affected, citing security sensitivity. It emphasized that it continuously monitors cybersecurity risks and implements protective measures when necessary.

Growing Global Caution Around Advanced AI Agents

Government institutions worldwide increasingly restrict AI tools due to similar concerns. Advanced AI agents can now perform multi-step actions such as reviewing emails, interacting with online services, and automating workflows. Tools like OpenClaw demonstrate how autonomous systems can execute complex sequences of tasks, raising questions about oversight and data exposure in sensitive environments.

The move also comes amid broader shifts in the AI landscape. OpenClaw’s creator recently joined OpenAI, while Microsoft is reducing its reliance on OpenAI by developing more in-house AI capabilities.

For now, the European Parliament’s decision signals a cautious approach toward AI adoption in government environments, prioritizing data protection and cybersecurity over rapid deployment of new AI-powered features.

Via Neowin