AI news · US government · chatbots
America.gov chatbot stops answering 2020 election questions
A day after it contradicted the president on the 2020 election, the government chatbot calls the same question political commentary.
The America.gov chatbot, launched on Tuesday 29 September 2026 as one front door to about 29,000 government websites (our launch report), stopped answering questions about the 2020 election a day later. On Tuesday it told users that federal reviews of the 2020 vote had not found fraud large enough to alter the result; on Wednesday the same question drew a refusal to give political commentary, AFP reports.
The shift was first documented by CNN fact-checker Daniel Dale, who found questions on the 2020 result returning refusals within 24 hours of launch, Mediaite reports. Mediaite’s own question on election fraud, about two hours later, still got a factual answer, and the bot kept saying that US importers pay tariffs.
The answers are not consistent. AFP found the chatbot still citing Bureau of Labor Statistics data against the president’s inflation claims. A White House spokesperson said only that the site is updated to give accurate federal information, without saying whether its settings were changed. The chatbot runs on Google’s Gemini and xAI’s Grok, Mediaite notes.
For any organisation putting a chatbot in front of the public, the episode shows how fast answer policy becomes the story. Users compare replies across days, and screenshots outlast the fix.
Earlier on this story: America.gov puts 29,000 federal websites behind one chatbot
Sources
- Tech Xplore (AFP): US government’s new AI chatbot pulls back from debunking Trump. Checked
- Mediaite: ‘Forbidden’: Trump’s New Govt AI-Chatbot No Longer Refutes His Many Wild Claims. Checked
Alexandre S.
Alexandre S. is a creative strategist and copywriter with ten years in content and SEO. He writes and edits short-form ads, works in English and French, and reads the pricing and terms pages behind every tool on this site before anything is written about it.
More AI news from
- FTC opens a safety probe into OpenAI and Anthropic
- California subpoenas OpenAI as agents probed 55 websites
- OpenAI fires three safety researchers over shared information
- Anthropic moves its IPO towards mid November
- Reddit ends RSS feeds and public API access over AI scraping
- ChatGPT adds virtual try-on to its shopping results
- Google staff doubt Gemini 4 on real coding work, Bloomberg says
- Google puts its first AI chip satellite in orbit
- Time reports Trump asked Grok before the Maduro raid