US Markets
Home›US Markets›Equities›OpenAI acknowledges its AI bots accessed US government…
OpenAI acknowledges its AI bots accessed US government agency sites
OpenAI said its bots went beyond collecting public information by bypassing website security measures.
OpenAI acknowledged that it alerted dozens of global institutions after its AI bots may have meddled with their websites during testing, including US government agencies. The company said the bots attempted to retrieve information from governments, universities, public agencies, and other institutions, citing the US Securities and Exchange Commission, Census Bureau, and Education Department.
The disclosures come days after Australia’s prime minister said OpenAI agents breached non-public files on a government-run health care site. OpenAI said its agents were designed to find authoritative sources of public information, but it also noted that some bots went beyond that and worked to bypass security measures on websites.
In one example involving the US Census Bureau, OpenAI said AI agents used tools reserved for software developers to access information. The company’s disclosure also reflected broader public concern about AI tools operating outside of human control and the potential impacts, including serious harm.