Sweden's government offices blocked ChatGPT and Claude today and kept Copilot, the one tool with a contract
From today, 26th Oct'26, staff in Sweden's Government Offices, which serve the prime minister and the ministries, can no longer open ChatGPT or Claude in the browser on their work computers. Microsoft Copilot stays as the only approved AI tool, after the office reviewed it and tightened its security settings. In a post on the staff intranet, reported by Sveriges Radio, the digitalisation director Mikael Wettercrantz said that using AI services without a formal agreement can have consequences for information security, and that there had been several examples of problematic use. Some staff can still reach OpenAI and Anthropic models through an internal platform the office built, and the Swedish Tax Agency and the National Board of Institutional Care already have similar blocks in place.
What it actually means
Read past the brand names and this is not a verdict on which model is better, because the same OpenAI and Anthropic models are still allowed through the office's own platform. What got blocked is staff typing government work into a consumer website with no contract behind it, where the office has no say over what happens to what they enter. Copilot passed because there is an agreement and settings the office controls.
Most companies have the same gap without the block, with people pasting client work into whichever chatbot they like on a personal account. If you run a team, the fix is the order Sweden chose: give people an approved route to good models before you shut the side doors, or they will find new ones. If you sell AI into public bodies or large firms, expect the first question to be about your contract and data terms rather than your benchmark scores.
Sweden did not ban the models; it banned using them without a contract.
Source: Sveriges Radio · Monday, 26 October 2026