Sweden’s Government Offices will block access to ChatGPT and Claude from employees’ work computers starting Monday. Mikael Grill Pettersson reports for Sveriges Radio Ekot that the move is intended to steer staff toward AI services approved by the government, including Microsoft Copilot.
The restriction reflects growing concern about employees entering sensitive information into public AI tools without formal agreements or security controls. In an internal message, Digitalization Director Mikael Wettercrantz says using AI services without contracts can have consequences for information security.
The Government Offices have also strengthened security settings for Copilot after a review, according to internal information seen by Ekot. Its press office did not provide details, citing security reasons.
Some employees will retain access to ChatGPT and Claude through a dedicated platform developed by the Government Offices. That platform provides controlled access to AI assistants. The organization had already limited use of the Chinese AI service DeepSeek.
Broader public-sector response
Other Swedish authorities have taken similar steps. The Swedish Tax Agency and the National Board of Institutional Care, known as SiS, have introduced blocks on some AI services. The Swedish Enforcement Authority is preparing its own decision.
Concerns extend beyond which tools staff use. Earlier this year, an internal audit at the Swedish Transport Administration found widespread use of roughly 10 AI services. The audit cited cases involving both sensitive information entered into open AI tools and the handling of AI-generated responses.
The actions show how public organizations are increasingly distinguishing between consumer AI tools and approved workplace systems. The central issue is not only access to generative AI, but also how information is protected, processed, and reviewed before it is used.
Stay up to date
AI for content creation: the latest tools, tips and trends. Every two weeks in your inbox:
