Swedish Government Restricts ChatGPT and Claude Over Data Privacy Risks on Staff Computers

Sweden’s Government Offices have officially blocked web access to popular AI chatbots ChatGPT and Claude on staff computers, citing severe data security and hidden IT risks.

Swedish Government Restricts ChatGPT and Claude

The sudden crackdown comes amid fears that sensitive public files could accidentally leak onto commercial servers abroad. To maintain productivity while ensuring data privacy, government employees have been directed to transition to Microsoft Copilot for their daily administrative tasks.

Unlike unregulated consumer web tools, this transition is backed by formal enterprise agreements. These contracts legally prevent tech companies from storing government queries or utilizing the data to train future AI algorithms. This level of compliance is critical for adhering to strict European privacy laws, which impose heavy penalties whenever sensitive workplace data is moved onto foreign servers located outside the region.

Sweden’s Actions

However, some select government workers can still utilize ChatGPT and Claude through a custom platform developed by a Stockholm firm. This localized system operates entirely on servers within Sweden. The Swedish Tax Agency has already implemented similar restrictions after discovering civil servants pasting confidential records into unapproved apps, and other public bodies are expected to follow suit.

Sweden’s actions reflect a growing global trend of securing digital borders against unregulated AI. The measures closely match recent AI bans enacted across various United States federal agencies. The European Parliament took an even tougher stance earlier this year, completely disabling AI features on work devices when it could not guarantee data safety.

Source