OpenAI is simultaneously expanding its education initiative to new countries while facing serious questions about ChatGPT's reliability in providing accurate information to voters. The company announced it is welcoming Singapore into its Education for Countries program, which now includes eight nations working to deploy AI tools in schools. However, a major study published the same week found that ChatGPT and other AI chatbots made significant errors when answering questions about recent elections in Scotland and Wales, spreading misinformation to millions of voters.

What Is OpenAI's Education for Countries Program?

OpenAI launched Education for Countries earlier this year at Davos as a government-led initiative to responsibly deploy AI in schools. The program brings together countries united by a shared goal of improving learning outcomes and unlocking economic opportunity through AI tools tailored for education.

The first cohort includes Estonia, Greece, Italy's CRUI, Slovakia, Trinidad and Tobago, Kazakhstan, the United Arab Emirates, and Jordan. Each country is advancing the program across three core pillars:

• Research-driven deployment: Using OpenAI's Learning Outcomes Measurement Suite, deployments begin as research partnerships that help governments and educators understand AI's impact on learners and build evidence on what works.
• Localized AI tools for learning: System-wide access to secure, compliant, and private ChatGPT, Codex, and OpenAI's API platform tailored specifically to teaching and learning needs.
• Teacher training and enablement: AI literacy, professional development, and certifications that help educators use these tools confidently and responsibly in classrooms.

Singapore's education system is among the world's best, and young Singaporeans are already highly active AI users. Around 43% of ChatGPT usage by 18 to 24 year olds is tied to learning and education. Through OpenAI for Singapore, the company is working with Singapore's Ministry of Education to support personalized learning, including helping students learn mother tongue languages more interactively.

How Are Early Participants Seeing Results?

The first cohort of countries is already reporting concrete signals of positive impact. Estonia's ChatGPT Edu deployment now reaches over 20,000 students and 4,600 teachers through the AI Leap Foundation. OpenAI is collaborating with AI Leap, the University of Tartu, and Stanford to understand how AI affects learning in real classroom settings, with findings committed to being shared publicly.

In Jordan, over 1 million students and over 100,000 teachers have engaged with an AI Education Assistant called Siraj as part of a broader national initiative. In Kazakhstan, a national ChatGPT Edu deployment across all 20 regions has seen over 84,000 educators complete AI-readiness training, with 9 in 10 educators surveyed saying ChatGPT Edu is useful for their work. In Slovakia, early university survey results show more than 9 in 10 educators report higher productivity, saving around 5 hours per week.

Why Are Election Officials Concerned About ChatGPT's Accuracy?

While OpenAI expands its education program, a major study by the thinktank Demos has raised urgent concerns about ChatGPT's reliability in providing accurate information to voters. The Electoral Commission has called for new legal controls over misinformation from AI chatbots after the research found serious mistakes during the recent Scottish election.

Demos ran a simulation before Scotland's May election by posing 75 questions to five free AI tools, including ChatGPT, Google Gemini, and Replika, about three real-life constituencies. The study found that AI services gave voters misinformation in response to 34% of the questions posed. ChatGPT, the most heavily used AI service, gave wrong information in 46% of its answers, including making up an expenses scandal, giving inaccurate replies on voter eligibility rules, and getting the election date wrong by two months.

The errors were widespread across different types of misinformation. AI tools variously invented fictitious scandals, gave the wrong date for the election, claimed wrongly that voters in Scottish elections needed ID at polling stations, and placed candidates in the wrong contests. Replika, a companion chatbot, performed the worst with errors in 56% of its answers. Google Gemini was wrong in 22% of cases, while Grok, linked to Elon Musk's social media platform X, had the lowest error rating at 9%.

An opinion poll of 2,005 British adults commissioned alongside the study found that 20% of voters had used AI chatbots or search tools to get information about the parliamentary elections in Scotland and Wales, and for English local councils, equivalent to 10 million people across the UK.

What Are Election Officials Demanding?

The Electoral Commission's chief executive, Vijay Rangarajan, has been pressing ministers to introduce legislation to make AI companies more accountable. He noted that voters want accurate information to help them engage with democracy, and it is concerning that AI tools have made the spread of false or misleading information dramatically faster and more accessible than ever.

"Voters want accurate information to help them engage with democracy and it is concerning that AI tools have made the spread of false or misleading information dramatically faster and more accessible than ever," said Vijay Rangarajan, chief executive of the Electoral Commission.

Vijay Rangarajan, Chief Executive, Electoral Commission

Rangarajan said ministers should introduce clearer duties on AI platforms to protect voters against misinformation and ensure algorithms do not mislead voters, particularly during critical election periods. This would give the media regulator, Ofcom, much clearer powers to enforce the law.

Azzurra Moores, an associate director at Demos, emphasized that this is a UK-wide concern, if not a global one. She noted that the accessibility of these AI tools, which are all developed and run by US corporations, is widespread in the UK, but the country does not yet have the legislative framework to protect the public from misinformation or democracy from its knock-on impact.

Moores suggested that ministers could quickly introduce legal requirements to make AI companies liable under UK defamation and electoral law, introduce mandatory safeguards on accuracy, and force AI firms to allow researchers to independently test how their internal data and training sets work.

What Did OpenAI Say About the Study?

OpenAI did not comment on the policy issues raised by Demos but argued that the research organization's approach was not typically how ChatGPT is used and appeared to be using an out-of-date version of the tool. The company noted that users are able to instruct ChatGPT to search the web for answers, which could improve accuracy.

The contrast between OpenAI's education expansion and the accuracy concerns highlighted by the Demos study underscores a fundamental tension in AI deployment. While the company is working with governments to responsibly integrate AI into classrooms with research partnerships and teacher training, the same tools are simultaneously spreading election misinformation to millions of voters who rely on them for civic information. The Electoral Commission's push for new legislation suggests that responsible AI deployment in education may require stronger guardrails and accountability measures across all public-facing AI applications.