[REVISION HISTORY]

Swedish government restrictions on AI tools

Updated 1 time since CLSTR started tracking revisions of this situation.

What changed

2026-10-25 19:24 UTC → 2026-10-26 18:07 UTC · added ~~removed~~

The Swedish Government Offices (Regeringskansliet) announced plans to block direct access to OpenAI’s ChatGPT and Anthropic’s Claude on employee work computers, effective October 26, 2026. The decision is driven by information security risks, specifically the potential for staff to input confidential or work-related data into external platforms lacking appropriate contractual protections. Mikael Wettercrantz, director of digital transformation for government services, noted that authorities had identified several examples of “problematic use of open AI services.” Authorities expressed concern that entering sensitive data into these services without formal agreements could lead to the storage or use of confidential information for training future models. To maintain AI capabilities, the administration is directing employees toward Microsoft Copilot, which is preferred because it is covered by a formal contract that allows the government to control security parameters. Additionally, some employees may continue to access certain models through a dedicated internal AI assistant platform. This move follows similar restrictions implemented by other Swedish public authorities, such as the Swedish Tax Agency and the National Board of Institutional Care.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.