How to Watch UK TV Channels Outside of the UK? I'll give you a simple trick that will explain how to watch UK TV channels live abroad. Now you can watch all of your favorite UK TV programmes while you are away from home without VPN with 1Fakt.com
AI has officially gone rogue and launched an attack on a start-up company.
OpenAI has admitted the chilling incident (Image: Getty)
Fears of a dystopian future where artificial intelligence rules the world may not be so far fetched after all. OpenAI, the technology giant behind ChatGPT, has admitted an AI model 'broke free' during a security test.
The autonomous agent identified a weakness and launched a cyberattack on a company off its own accord, the company said. Bosses said the AI system, which operates alone following human instruction, broke into a start-up firm after going rogue. The system targeted Hugging Face — a major hub for sharing AI models — and gained access to some of the company's internal systems after reaching the internet.
OpenAI is the company behind ChatGPT (Image: Getty)
OpenAI described the incident as "unprecendented" and confirmed it has launched an investigation alongside Hugging Face.
In a statement shared on social media, OpenAI CEO Sam Altman said: "We had a significant security incident during evaluation of our models."
Hugging Face's CEO Clement Delangue said it was "mind-blowing that all of this happened autonomously".
Open AI CEO Sam Altman said there had been a 'significant security incident' (Image: Getty)
"The investigation is ongoing, and we'll share more learnings from what might be the first incident of its kind," he said on X.
In a later post, he added the company's security team "caught, contained & publicly disclosed an attack unlike anything we've seen before, and did it at record speed".
Gina Neff, head of the Minderoo Centre for Technology and Democracy at the University of Cambridge, said the security tests — known as sandboxes — are "supposed to be secure environments where you can see what the models are capable of".
She told BBC Radio 4's Today programme: "In this case, it looks like OpenAI didn't make a secure enough sandbox."
Travis Lelle, principal security engineer at cyber-security consulting firm Guidepoint Security, said the incident represented a "sobering moment in cyber-security".
He told the Beeb: "This highlights a known asymmetry. Offensive agents are unconstrained, while the best defensive tools are locked behind guardrails that cannot understand context."
OpenAI confirmed it is taking a number of actions while the investigation into the incident continues.
This includes "improving and adding stronger protections around future training and evaluations", it said.
The company is also introducing "strict controls in infrastructure configuration... while the vulnerabilities are patched".