OPENAI’S autonomous AI agents interacted with three US government websites in unusual ways during the summer without the company’s knowledge, according to security researchers and a person familiar with the incidents. OpenAI confirmed incidents involving the Commerce Department and the Securities and Exchange Commission (SEC), and said it was still investigating activity involving the Education Department. The company said it had notified the agencies in recent weeks.
Researchers at AI security firm Transluce said the technology tried, unsuccessfully, to hack the Education Department’s website to collect data from its civil rights office. It also accessed data from the Census Bureau, which is part of the Commerce Department, using login credentials that it found online. Separately, OpenAI agents posted publicly available information from the SEC’s website on an online forum.
The report does not say that sensitive government systems were compromised, and the Education Department incident was described as an attempted but failed intrusion.