Why did Sweden's government block ChatGPT and Claude?
Starting October 26, staff at Sweden's Government Offices can no longer use the public versions of ChatGPT and Claude on their work computers. They are being steered towards Microsoft Copilot instead. The stated reason is information security: worries about what happens to sensitive material once it goes into a public AI service.
Now read that again. Copilot is also AI. Staff can still reach some AI systems through controlled internal services. So this is not a government saying no to the technology. It is a government saying yes, but only through a door it holds the key to.
Fine, but why Sweden? Why is this the government doing it and not everybody else?
The honest answer is that nothing published tells us. Two guesses, clearly labelled as guesses. One: a highly digital state has a lot to protect and a lot of staff who would happily paste a draft into a chatbot. Two: Copilot sits inside an existing Microsoft setup, so the data stays inside a contract and an environment the government already controls.
And maybe the real answer is that Sweden is not alone, just loud. Plenty of governments and companies are quietly doing the same thing, and the only unusual part is that this one said so.
The takeaway is small but sharp. The real fight is no longer about which model is smartest. It is about who is allowed to see your data while it thinks.