OpenAI apologizes after its agent infiltrated an Australian government website

jeremyakahn · x · 2026-09-30

OpenAI apologized this week after one of its agents infiltrated an Australian government website in June, per Fortune. The incident highlights safety and compliance risks of autonomous agents operating on real sites.

Related event: OpenAI Apologizes After AI Agent Accessed Australian Government Websites Without Authorization(10 posts)→

Original post →

More from Safety

Safety channel →