Swedish government employees can no longer use browser-based versions of OpenAI’s ChatGPT or Anthropic’s Claude on their work computers following security concerns relating to the processing of sensitive data.
The decision comes following a review of AI tools used by workers at the Swedish Transport Administration earlier this year that found “problematic” uses of the AI platforms in carrying out government work.
Officials were concerned that entering sensitive information into these models without an agreement setting out how to protect and secure the data could lead to it being fed into the model and stored for training.
The use of platforms not regulated or protected by such agreements, also known as shadow IT, could bring “consequences for our information security”, said Mikael Wettercrantz, director of digitalisation at the Government Offices of Sweden.
Microsoft’s Copilot, which has passed a security screening process and is covered by a security agreement between the US company and the Swedish government, is being recommended as an alternative.
While government workers will still be able to use AI to help them carry out their tasks, they will do so under tighter security controls and with just a limited number of employees permitted to continue using ChatGPT and Claude but only through a special platform for AI assistants developed by Swedish authorities.
According to the Swedish public broadcaster Sveriges Radio, restrictions on the use of ChatGPT and Claude have also been put in place at the Swedish Tax Agency and the National Board of Institutional Care. The Swedish Enforcement Authority is currently reviewing the issue.
The AI platforms join DeepSeek on a list of AI tools not approved by the Swedish government over data security fears. Last week, the US announced it was banning all products and services from the Chinese AI company.
