How Governments Like Sweden Are Using ChatGPT and AI Chatbots in Public Services

Governments worldwide are increasingly integrating artificial intelligence (AI) chatbots like ChatGPT into public services to boost efficiency, streamline operations, and enhance citizen engagement. In Sweden, a global leader in digital governance, the use of AI tools, including OpenAI’s ChatGPT, has sparked both innovation and controversy, highlighting the opportunities and risks of AI in public administration.

Sweden’s Experiment with AI in Governance

Swedish Prime Minister Ulf Kristersson recently made headlines by admitting he uses ChatGPT and Mistral’s LeChat to seek “second opinions” on policy matters, as reported by Dagens Industri on August 5, 2025. Kristersson described using these tools to brainstorm ideas, such as exploring alternative approaches or reviewing what other countries have done, emphasizing that they serve as supplementary aids rather than decision-makers. “I use it myself quite often, if for nothing else than for a second opinion,” he said. His colleagues in the Moderate Party-led coalition also employ AI for daily tasks, reflecting a growing trend in Swedish governance.

However, Kristensen’s openness triggered backlash. Critics, including tech experts and media outlets like Aftonbladet, accused him of succumbing to “the oligarchs’ AI psychosis,” arguing that reliance on AI risks introducing bias or misinformation into policy decisions. Virginia Dignum, a professor of responsible AI at Umeå University, warned, “The more he relies on AI for simple things, the bigger the risk of overconfidence in the system. It is a slippery slope.” She emphasized that AI cannot provide objective political opinions, as it reflects the biases of its training data.

This isn’t Sweden’s first brush with AI controversy. In 2024, Social Democrat MP Olle Thorell used ChatGPT to generate 180 parliamentary questions, drawing criticism for overburdening ministers’ staff. Despite the pushback, some Swedish officials defend AI’s role. Mathias Sundin, a member of Sweden’s AI Commission, argued that Kristensen’s approach is pragmatic, leveraging AI’s vast knowledge base—trained on “the entire internet and hundreds of thousands of books”—as a valuable reference tool.

AI in Swedish Public Services

Beyond high-level policy, Sweden is deploying AI chatbots across public services to enhance efficiency. The Swedish Tax Agency has implemented an AI-powered chatbot to handle routine taxpayer inquiries, reducing call center wait times by 30% since its launch in 2023. Similarly, the City of Stockholm uses a chatbot to assist residents with municipal services, such as parking permits and waste management queries, processing over 10,000 requests monthly. These tools, often built on custom AI models rather than ChatGPT, prioritize data privacy and compliance with GDPR.

Sweden’s healthcare sector is also experimenting with AI. The Västra Götaland region piloted a chatbot in 2024 to triage non-emergency medical inquiries, directing patients to appropriate care and freeing up staff for critical cases. The system, developed with local AI firm AI Sweden, reported a 25% reduction in unnecessary clinic visits. However, concerns about data security and the handling of sensitive health information remain, with experts like Simone Fischer-Hübner of Karlstad University cautioning against using commercial tools like ChatGPT for sensitive tasks due to potential data leaks to U.S. servers.

Global Trends in AI-Powered Public Services

Sweden’s approach mirrors a global shift toward AI in governance. In the United States, OpenAI announced a partnership on August 5, 2025, to provide ChatGPT Enterprise to the federal workforce for $1 per year, following a successful pilot in Pennsylvania that saved public servants an average of 95 minutes daily on routine tasks like document drafting and data analysis. The initiative, part of the Trump administration’s AI Action Plan, includes “strong guardrails” to ensure transparency and alignment with public missions.

In the UK, the Department for Science, Innovation and Technology has integrated AI chatbots to streamline administrative tasks, though Secretary Peter Kyle faced criticism in 2024 for using ChatGPT to analyze business AI adoption and select podcast appearances. Scotland’s MP Graham Leadbitter openly uses AI to draft speeches, calling it a “good basis to work from” while retaining final control.

Singapore has taken a more structured approach, with its Smart Nation initiative deploying AI chatbots across government services, from tax filing to citizen feedback. The country’s GovTech agency reported a 40% increase in query resolution speed since adopting AI in 2023. Similarly, Estonia, a digital governance pioneer, uses AI chatbots to assist with e-residency applications, handling over 50,000 inquiries annually with 90% accuracy.

Challenges and the Path Forward

While AI chatbots offer efficiency, their use in public services raises ethical and practical concerns. In Sweden, public sentiment on X reflects unease, with users echoing Dignum’s sentiment: “We didn’t vote for ChatGPT.” Critics argue that AI’s potential for “hallucinations”—generating plausible but incorrect information—could undermine trust in government decisions. Aftonbladet’s Signe Krantz wrote, “Chatbots would rather write what they think you want than what you need to hear,” highlighting the risk of sycophantic outputs.

Data privacy is another hurdle. Jakob Ohlsson, an AI consultant, criticized Kristensen’s use of ChatGPT, noting that even non-sensitive queries could reveal strategic thinking when aggregated, potentially compromising national interests on U.S.-based servers. The EU’s AI Act, effective August 1, 2025, mandates transparency and bias mitigation, prompting calls for Sweden to align with these standards by establishing clear guidelines for AI use in governance.

Despite these challenges, proponents see AI as inevitable in public services. Elias Rosell, a Swedish political analyst, predicted that AI will become “one of the key factors shaping this century,” urging governments to balance innovation with accountability. As Sweden and other nations expand AI adoption, the focus is shifting toward robust oversight, local AI solutions, and public trust to ensure chatbots enhance, rather than undermine, democratic processes.