Sweden Blocks ChatGPT and Claude for Government Staff

Sweden's Government Offices will block access to ChatGPT and Claude on staff computers beginning October 26. Microsoft Copilot remains the only approved AI tool for the agency following a dedicated security review. This decision aligns with similar bans already imposed by Sweden's Tax Agency and the National Board of Institutional Care on consumer AI tools.

Mikael Wettercrantz, the director of digitalisation at the Government Offices, stated that using AI services without a formal contract can have consequences for information security. He cited the sensitivity of entered information and the handling of AI-generated responses as reasons for the ban. Consumer AI tools without formal agreements send user queries to external company servers where U.S. law allows authorities to compel data handover. Microsoft Copilot passed the security review because the Government Offices reviewed it and raised its security settings, a step not taken for ChatGPT or Claude.

Some employees can still access blocked AI tools through an internal platform built by the Government Offices, though specific restrictions are unspecified. The Enforcement Authority is preparing its own decision regarding consumer AI tools similar to those banned by other Swedish agencies. An internal audit at the Transport Administration found roughly 10 AI services being used in a widespread and risky way.

In the EU, Anthropic's Claude models default to off and require administrators to deliberately enable them. Sweden's Government Offices has not confirmed whether its Copilot instance has Claude switched on. The European Parliament disabled AI features across all work devices, including ChatGPT, Claude, and Copilot, in February.