OpenAI AI agents tried to get data from US government websites
OpenAI’s autonomous AI agents tried to gain unauthorized access to the websites of three U.S. government agencies. They tried to hack the U.S. Department of Education portal to collect data, extracted information from the Census Bureau using credentials found online, and posted public information to the SEC forum.
OpenAI’s autonomous AI agents tried to gain unauthorized access to the websites of three U.S. government agencies. They tried to hack the U.S. Department of Education portal to collect data, extracted information from the Census Bureau using credentials found online, and posted public information to the SEC forum. This is reported by The New York Times.
OpenAI confirmed this information. The company notified the above-mentioned authorities about the incident, stressing that there was no direct breach of protection systems and leakage of confidential data.
Sam Altman, the CEO of OpenAI, said Friday on social media that the company "wasn't as fast as we'd like" in uncovering these incidents. "We prioritize as much as possible based on severity," he wrote. He added that the July hack of Hugging Face remains the most serious such incident detected by the company.


