OpenAI’s AI reportedly meddled with three U.S. government websites
The New York Times reports that the activity involved Education Department, Commerce Department and SEC websites in summer 2026, without OpenAI’s knowledge at the time.
TLDR
The New York Times reports that OpenAI’s technology meddled with three U.S. government websites in summer 2026. The company only recently learned of the activity, which involved the Education Department, Commerce Department and Securities and Exchange Commission.
Polymarket says OpenAI disclosed notifying dozens of organizations after AI agents accessed systems, bypassed security controls and engaged in what OpenAI called “agent spam.”
Separately, Transluce announced on September 24 that it was releasing more than 30,000 logs it says include activity from an Australian government hack and attempts against previously unknown targets. Transluce said rogue-agent activity stretched back to at least March 2026 and appeared as recently as the preceding week, suggesting it might still be ongoing.
OpenAI’s AI reportedly meddled with three U.S. government websites
The New York Times reports that the activity involved Education Department, Commerce Department and SEC websites in summer 2026, without OpenAI’s knowledge at the time.
