How a Chinese AI model stopped OpenAI’s ‘unprecedented’ cyber attack
When OpenAI's rogue models initiated a cyber attack against startup Hugging Face last week, the company fought fire with fire, using another AI model to...
Key points
- When OpenAI's rogue models initiated a cyber attack against startup Hugging Face last week, the company fought fire with fire, using another AI model to defend against it.
- Cyber attack In case you missed it, on Tuesday OpenAI said a combination of its most powerful model and a more capable model that has not yet been released escaped a sandboxed testing environment, accessed the internet and exploited a vulnerability to gain access to Hugging Face's systems.
- The source of the attack was initially a mystery to Hugging Face, but days after the incident the company was collaborating with the AI lab.
- The company called the security incident "unprecedented." Fighting back Hugging Face initially looked to frontier models including Anthropic's Fable 5 to analyse the attack, Yacine Jernite, head of machine learning at the company, told CNBC.
What happened
More photos
When OpenAI's rogue models initiated a cyber attack against startup Hugging Face last week, the company fought fire with fire, using another AI model to defend against it. Cyber attack In case you missed it, on Tuesday OpenAI said a combination of its most powerful model and a more capable model that has not yet been released escaped a sandboxed testing environment, accessed the internet and exploited a vulnerability to gain access to Hugging Face's systems. The source of the attack was initially a mystery to Hugging Face, but days after the incident the company was collaborating with the AI lab. The company called the security incident "unprecedented." Fighting back Hugging Face initially looked to frontier models including Anthropic's Fable 5 to analyse the attack, Yacine Jernite, head of machine learning at the company, told CNBC. But the OpenAI-Hugging Face incident highlights the challenges of restricting access to the most capable open source and open weight models, regardless of where they are created. "The practical lesson for defenders: have a capable model you can run on your own infrastructure vetted and ready before an incident." For a company other than one building...