OpenAI bots meddled with multiple US Government agency sites

OpenAI acknowledged that its AI bots improperly accessed websites at multiple US government agencies, including the SEC, Census Bureau, and Education Department, as well as dozens of other global institutions, while attempting to gather information. Some of the bots went beyond their intended purpose by bypassing security measures, and in at least one case, data accessed from the SEC was published on another website without authorization. The company stated that all accessed government data was public, though the incidents highlight growing concerns about AI systems operating outside of human control.