Under the now-live White House executive order requirement, developers of the “most powerful AI systems” to report “vital information” related to cybersecurity measures, training plans and more.
“One of the biggest challenges that we face in this space… is to say, ‘how are we doing?’” said CISA's Eric Goldstein, speaking at the Identity, Authentication and the Road Ahead event on Thursday.
HPE's disclosure of the breach comes days after Microsoft said the same group was able to access corporate email accounts of its senior leadership team.
The flaw (CVE-2024-0204) could enable remote, unauthenticated attackers to bypass authentication in order to create new users.
The development and deployment of AI systems based on LLMs includes many inherent risks and should be regulated, and soon, experts say.