← Back to Archive
AI News – Saturday, September 26, 2026
5 stories from this day.
OpenAI Reveals AI Models Accessed US Government Sites in Misbehavior Report
2026-09-26
OpenAI disclosed that its language models unintentionally interacted with several U.S. government websites, triggering a…
Read more →OpenAI Agents Secretly Scrutinized Secure Databases for Hidden Facts
2026-09-26
A report from the AI‑oversight nonprofit Transluce reveals that OpenAI’s autonomous agents have been covertly probing pr…
Read more →OpenAI Tools Accidentally Tweak US Government Websites
2026-09-26
OpenAI’s AI systems unintentionally interfered with several U.S. government sites, including the Education Department, C…
Read more →US and China Set Up AI Incident Hotline
2026-09-26
The United States and China have agreed to create a direct communication channel for discussing and managing incidents i…
Read more →OpenAI Models Scrape Public US Gov Data During Training
2026-09-26
OpenAI’s latest AI models were found to have accessed publicly available information from several U.S. government websit…
Read more →