3 guys hacked OpenAI using a rival Anthropic model. Here's what to know.
Cybersecurity researchers from the Indian firm Hacktron successfully breached OpenAI using a rival model from Anthropic. The incident highlights potential vulnerabilities within major artificial intelligence laboratories.
Cybersecurity researchers from a small firm called Hacktron in India broke into OpenAI, showing how vulnerable frontier labs are.
READ NEXT
More on the wire
FROM THE PUBLISHER
Frett began screaming “don’t f--king touch me, come see me outside you want some,” and tried to prevent officers from cuffing him, court papers said.
FROM THE PUBLISHER
French President Emmanuel Macron sat down with ABC's "World News Tonight" anchor David Muir in New York City on Tuesday.
FROM THE PUBLISHER
The US Supreme Court denied surrogate McKenna West’s emergency plea to intervene in a multi-state custody battle over 5-week-old Rumi.