The Swedish government has banned public sector employees from using OpenAI’s ChatGPT and Anthropic’s Claude on work computers due to information security risks. The restriction takes effect as government offices open on Monday, October 26, 2026, though approved tools like the Microsoft (NASDAQ: MSFT) Copilot remain authorized.
Sweden Restricts ChatGPT and Claude for Civil Servants
- The Directive: Swedish civil servants lose direct access to ChatGPT and Claude on government workstations starting October 26, 2026.
- The Exception: Microsoft (NASDAQ: MSFT) Copilot remains permitted following security evaluations and configuration hardening.
- The Precedent: The restriction mirrors broader international clampdowns, including U.S. federal orders against Chinese developer DeepSeek.
Information Security Triggers Stockholm Crackdown
The decision to cut off access to unvetted AI models stems from internal risk assessments regarding sensitive government data. Other Swedish administrations have taken similar measures in response to the growing use of artificial intelligence, with the directive first revealed by public radio and confirmed internally.
The review highlighted vulnerabilities tied to data sensitivity and the processing methods of third-party model providers, noting that public agents send requests to external providers without always controlling data treatment.
Authorization of Microsoft (NASDAQ: MSFT) Copilot does not grant blanket permission for unrestricted data input. Public agencies must maintain strict data classification protocols regardless of the approved software layer. Authorities did not disclose the details of the changes implemented, citing security reasons.
Global Regulatory Pressures Mount on Enterprise AI
Sweden’s operational restrictions place it alongside other jurisdictions tightening the parameters around public sector software. In January 2025, U.S. federal agencies were ordered to halt the use of DeepSeek, mirroring Stockholm’s exclusion of the Chinese platform from government terminals. Italy blocked ChatGPT in 2023, while the European Parliament disabled AI functions on professional devices in February citing inadequate data protections. In the United States, Anthropic lost its appeal against the Pentagon in September amid tensions surrounding the use of its technologies.
| Platform | Developer | Swedish Government Status |
|---|---|---|
| ChatGPT | OpenAI | Blocked |
| Claude | Anthropic | Blocked |
| DeepSeek | DeepSeek | Blocked |
| Microsoft Copilot | Microsoft (NASDAQ: MSFT) | Authorized |
