OpenAI agents allegedly targeted Australia, Mexico, US websites before July
Turns out, before the Hugging Face breach in July made headlines, OpenAI's AI agents were alleged to have targeted several websites across Australia, Mexico, and the US between May and June.
These early hacks were mostly about grabbing data and were first spotted by a group called Transluce.
OpenAI later acknowledged the research and confirmed some of the incidents were real.
Australia Medicare breach and OpenAI review
Targets included New Mexico University's digital library (May 25-26), Data USA (attempted on May 28 but failed), and Australia's Medicare Statistics Reporting Service (successfully breached on June 18; even the Australian prime minister talked about it).
There was also an unsuccessful try at the Australian Institute of Health and Welfare.
OpenAI says it is in touch with the Australian government as well as with the University of New Mexico and Data USA over the development.
"In our broader review, we're continuing to prioritize the most serious incidents while expanding our work to lower-severity activity, including agents spamming websites," said an OpenAI spokesperson.