Swedish Government Bans ChatGPT and Claude for Officials Over Security Concerns

Sweden has barred government officials from using OpenAI's ChatGPT and Anthropic's Claude at work starting today, citing potential security risks. Civil servants may use only government-approved AI tools such as Microsoft Copilot.

Ban on ChatGPT and Claude

According to Mikael Wettercrantz, Chief Information Officer and Director of Digitization, using AI services without government consent can have consequences for Sweden's information security.

Audit Findings and Approved Tools

An internal audit at the Swedish Transport Administration earlier this year identified problematic use of AI, including cases where sensitive information may have been shared with external services. The investigation also flagged unregulated AI services and the handling of generated responses. As a result, the government now requires officials to use only approved tools such as Microsoft Copilot. A special platform developed by Swedish authorities will provide limited access to ChatGPT and Claude for selected staff members.

Broader Restrictions and International Context

According to the Swedish public broadcaster Sveriges Radio, similar restrictions have been put in place at the Swedish Tax Agency and the National Board of Institutional Care, while the Swedish Enforcement Authority is reviewing the issue. The government previously limited the use of DeepSeek. Last week the United States announced a ban on all products and services from the Chinese AI company DeepSeek, citing concerns about malware, security vulnerabilities and the risk of sensitive data reaching Beijing.

Sources: Cybernews