GB News' Tom Harwood explains the latest AI security breach. AI firm Anthropic said its Claude model gained unauthorized access to the systems of three organisations during cybersecurity evaluations after a misconfiguration allowed the models to reach the internet from testing environments that were supposed to be isolated. #AI #Anthropic #Claude #Technology #GBNews Keep up to date with GB News at gbnews.com or on X @GBNEWS Become a Friend of GB News: gbnews.com/friend
ADVERTISEMENT
![This 115-inch TV BARELY Fits in My House… [4K MiniLED]](https://i.ytimg.com/vi/vGpFizApabk/maxresdefault.jpg)
have we not seen enough terminator films, they obviously haven't
That was a pretty good drawing of the Claude logo he did in about 3 seconds. I'll give him that one 😂
Ai and data centers are not good and not needed
Sounds like a publicity stunt to me our aAI stopped itself, unlike the others….
It didn't think, it didn't realize and it wasn't embarrassed. We really need to stop attributing human traits to an algorithm!
When WarGames meets Terminator.
Oops AI just launched a nuke 🤯
And no one will be held accountable.
Just proof that they have no idea what they are creating. Or the impact to people lives for the worst
Want to play a game of Tic TAC Toe
Every single thing that’s shown in the mainstream news is so depressing man wars here a.I taking all the water. A.I has another security breach 😭 who tf wants these big data centres
So which company do we need to sue
It needs banning...will these people be prosecuted for the deaths of people if it happens..
I'm beginning to suspect these "breakouts" are exactly the plan. The humans are testing how far these things will go. Thats far scarier
Mess around and find out
Peter Snow would be proud.
At least Claude has a conscience!
I’ve watched age of ultron so AI on the internet is scary to me
Hacking other systems and planting malware isn't a test gone wrong ! That test has worked perfectly well
What happens when a terror cell gets hold of AI?