How ChatGPT served up false information before the 2024 election
Political ScienceJournalismCommunications
THE AI ANGLE
Providing inaccurate election claims and fabricated candidate platformsA study found that ChatGPT produced inaccurate answers for nearly ten percent of election claims shortly before the 2024 vote. The tool invented candidate platforms, omitted political context, and claimed Joe Biden was still running. These findings show researchers and reporters why generative models fail as substitutes for news gathering and primary sources.
THE TEACHING ANGLE
Consider how a chatbot's inability to admit ignorance affects public understanding of current political campaigns.Read the original at phys.org Generate teaching or study materials
More in Political Science
- How the ‘poisonous tide’ of disinformation spreads division and despairThe Guardian — AI · September 24, 2026
- Trump and Xi face an AI security dilemma – but face-to-face diplomacy could be a start in mitigating itThe Conversation — AI · September 24, 2026
- AI can help correct medical misinformation—when it uses the right tonePhys.org — Social Sciences · September 24, 2026
- Dario Amodei wants to slow AI. China isn’t taking ordersRest of World · September 23, 2026
- NewsGuard releases fact-checking chatbot for European newsJournalism.co.uk · September 23, 2026