Chatbots struggle with news accuracy and sourcing ahead of US midterms
A study by Forum AI reveals that major chatbots struggle with accuracy and bias in their responses regarding elections and geopolitics. The research found that these chatbots provided inaccurate or biased answers 90% of the time, particularly in election-related queries. Concerns are growing as the midterm elections approach, highlighting the need for accountability in AI-generated news.
- ▪Four major AI chatbots were tested for accuracy in answering questions about news topics.
- ▪The study found that 90% of the chatbot responses about elections were inaccurate or biased.
- ▪Grok was the worst offender, returning factual errors nearly 52% of the time.
Opening excerpt (first ~120 words) tap to expand
Chatbots struggle with news accuracy and sourcing ahead of US midtermsSign up now: Get ST's newsletters delivered to your inboxWhen ChatGPT, Claude and Gemini returned biased answers, they aligned with the political left, and Grok primarily tilted in favour of the political right.PHOTO: AFPPublished May 20, 2026, 10:22 PMUpdated May 20, 2026, 10:22 PMSAN FRANCISCO – Four major artificial intelligence chatbots – OpenAI’s ChatGPT, Alphabet Inc’s Google Gemini, Anthropic’s Claude and xAI’s Grok – are struggling to fairly and accurately answer questions about elections and geopolitics, according to a new study from Forum AI.Researchers asked the four chatbots more than 3,100 questions about a wide range of news topics, like politics, healthcare and foreign affairs.They found that the…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Straits Times — World.