TLDR: Swedish Prime Minister Ulf Kristersson has ignited a public debate after revealing his frequent use of AI chatbots, including ChatGPT and LeChat, to inform policy decisions. While Kristersson defends his use for gaining ‘second opinions’ and exploring different perspectives, experts and critics raise significant concerns about data security, the objectivity of AI, and the implications for democratic accountability.
Swedish Prime Minister Ulf Kristersson has sparked a significant national discussion following his candid admission of regularly employing artificial intelligence chatbots, such as OpenAI’s ChatGPT and the French messaging app LeChat, to assist in his policy decision-making process. The revelation, made in an interview with Dagens Industri published on August 3, 2025, has drawn mixed reactions across Sweden, prompting a broader debate on the role of AI in governance.
Kristersson, the 61-year-old leader of the center-right Moderate Party, stated that he uses these AI tools ‘quite often,’ primarily for a ‘second opinion’ or to explore alternative viewpoints. ‘What have others done? And should we think the complete opposite? Those types of questions,’ Kristersson explained, likening his approach to a doctor seeking multiple opinions before a diagnosis. He emphasized that no security-sensitive information is uploaded to these platforms, and their use is more akin to ‘a ballpark’ for initial insights. The Prime Minister’s Office spokesperson, Tom Samuelsson, reiterated that Kristersson exercises caution in his AI usage.
However, this admission has not been without its critics. Concerns have been voiced by tech experts and political commentators regarding the potential risks and implications of relying on AI for serious political matters. Simone Fischer-Hübner, a computer science researcher at Karlstad University, cautioned about inherent security risks associated with politicians using platforms like ChatGPT. Virginia Dignum, a professor of AI ethics at Umeå University, argued that AI, by its nature, cannot offer truly objective political opinions, warning of a ‘slippery slope’ where overconfidence in AI systems could undermine reliability. ‘We must demand that reliability can be guaranteed. We didn’t vote for ChatGPT,’ Dignum stated, encapsulating a sentiment echoed by many.
Further criticism came from figures like Jakob Ohlsson, an AI enthusiast and consultant, who described Kristersson’s approach as ‘amateurish.’ Ohlsson highlighted the risks of inputting political thoughts into a language model that the user does not fully understand, owned by a company outside of governmental control, with servers potentially located in countries with uncertain democratic futures. Writer and lecturer Signe Krantz, in Aftonbladet, questioned why the Prime Minister would consult ‘random number generators’ instead of his ‘large and well-paid staff of experts,’ suggesting that fixing AI’s mistakes could take longer than original human effort. Krantz’s viral critique, ‘We didn’t vote for ChatGPT,’ has become a rallying cry for those wary of AI’s increasing influence in governance.
Also Read:
- AI Expert Predicts Potential for AI-Driven Governance, Even a Presidential Role, by 2032
- Bahrain Launches Comprehensive National AI Policy for Ethical Digital Advancement
Despite the backlash, Kristersson maintains that his government aims to embrace AI and digitalization to efficiently process vast amounts of data within national administration, believing technology can be highly useful for experts and advisers. He also noted that other cabinet members are integrating AI into their daily work. This ongoing debate underscores the complex challenges and ethical considerations as artificial intelligence increasingly intersects with the highest levels of political decision-making.
