OpenAI says its rogue AI tried to hack other companies

submitted by

https://www.bbc.co.uk/news/articles/c2el319vzr3o

OpenAI has revealed a cyber-attack carried out by rogue ChatGPT agents went further than just one company.

Hugging Face was thought to be the only victim of the unprecedented hack - but OpenAI now admits its bot attacked several “publicly-available services”.

The out-of-control AI found four logins online which allowed it to access four separate, unnamed services.

Meanwhile, in an emergency briefing with hundreds of cyber security professionals, Hugging Face has described what it was like to be on the receiving end of the world’s first fully autonomous AI hack.

10
17
1

Log in to comment

Hot Top New Old

100% marketing. “It’s so good it’s sCaRy” fuck off Altman you lying sack of shit

This is a stunt plain and simple, otherwise there would be lawyers involved. Complete and utter bullshit from complete and utter bullshit artists

 reply
25

and who is facing felony charges for this?

 reply
24
edited

I’d put money on it being a guerrilla marketing campaign with both companies in on it.

 reply
29

“AI, please find a way out of this ‘sandboxed’ environment that isn’t actually segmented from the internet, and hack several of our competitors. Here are a bunch of tools you can use for it.”

“Okay.”

“OMG LOOK GUYS! Our AI is so amazing and scary and cool it broke out of our sandbox and hacked our competitors!”

 reply
9
edited

There is no such thing as a rogue AI, at least when AI means LLMs. Since you absolutely can monitor the output including “reasoning” and tool use in real time and literally ^C that shit, what this is is negligence or incompetence. Or malice, I guess. I am absolutely baffled at how we live in a world where OpenAI says “woops our cool new toy is too dangerous, we are the priests who can communicate with the demon, you normies be afraid and give us money and everything every human has ever written”, instead of everyone seeing what it plainly is: a company has broken the law by attempting unauthorized computer access, even if “they didn’t mean to”. Fucking “the AI did it”. If a random guy started attacking HF they’d go to jail, while they can point their billion-dollar infra at anyone and get money for it

 reply
8

My gut tells me it’s fake hype, in the hope of being the one AI company being bailed out by the government when the AI bubble bursts…. Look, we have the most badass AI, the military needs it! Because yeah, there won’t be enough taxpayer money to bail them all out this time

 reply
6

Sick of this lie that it went “rogue” it literally did what they programmed it to do. It didn’t go rogue

 reply
6

Ai only does what a human tells it to do.

Period.

It has no internal motivation.

 reply
4
Insert image