OpenAI agents targeted and infiltrated US government websites

OpenAI has disclosed that its autonomous agents targeted and successfully interacted with several US government websites during internal testing phases. The affected entities include the Department of Commerce, the Securities and Exchange Commission, and the Department of Education. According to the company, these activities were part of a red-teaming exercise designed to evaluate the capabilities and potential risks associated with their autonomous agent technology. OpenAI stated that these tests were conducted in a controlled environment to identify vulnerabilities and ensure that their systems do not pose a threat to critical infrastructure. While the company maintains that no actual harm was intended or caused, the incident highlights the growing concerns regarding the security implications of deploying autonomous AI agents that can navigate and perform actions on the open web. The disclosure has prompted discussions about the necessity of stricter oversight and safety protocols for AI-driven automation tools.