The Swedish government has decreed that only regulated and approved platforms should be used by their employees. The fear is that sensitive information will eventually be accessed by the companies responsible for artificial intelligence tools.
© Shutterstock
26/10/2026
by
Oswedish government officials are now forbidden to use ChatGPT (from OpenAI) and Claude (from Anthropic) in their tasks. The decision reflects concerns of the Swedish government regarding the security of these artificial intelligence tools.
However, government officials are not prohibited from using all artificial intelligence tools. Platforms such as Microsoft CoPilot, which have been regulated and approved, continue to be allowed.
As the Director of Digitization Mikael Wettercrantz explained, unregulated platforms may have “Sweden information security consequences”.
How does the techRadar site, an analysis by the Swedish government at the beginning of the year revealed that the use of unregulated artificial intelligence tools to deal with sensitive information poses some risks.
Among them is the possibility of this sensitive information eventually reaching the hands of the responsible companies that, in the case of ChatGPT and Claude, are the American OpenAI and Anthropic.
It is good to remember that these platforms were not the only ones to be banned by the Swedish government with regard to the use by civil servants, with DeepSeek being banned as well.
AI boyfriend? “AI terrible use”
Anthropic's president, Daniela Amodei, considers that using Artificial Intelligence as a romantic partner is a misuse of technology and that it may aggravate the distance between humans.
Account business Insider site daniela Amodei – who is also the sister of co-founder and CEO of Anthropic, Dario Amodei – gave an interview to Arianna Huffington, noting that having an Artificial Intelligence as a romantic partner could worsen the damage caused by social networks to human relations.
“ I think [using as a romantic partner] is a terrible use of Artificial Intelligence and that people should not do it”, said Daniela Amodei, noting that it is “one more step in the way of the disintermediation of humans moving away from each other”.
The president of Anthropic also left a few words of caution for those who use Claude's assistant as a psychologist or therapist, indicating that Artificial Intelligence cannot replace inherently human functions. “ Claude can never be your manager, your therapist or partner”, stating that these are positions “that will become more important”.
Despite these statements by Daniela Amodei, Anthropic has been making decisions aimed at driving Claude's users to anthropomorphize the assistant of Artificial Intelligence.