Maria Bartiromo discusses OpenAI's AI testing agent that has broken ground and hacked the AI platform.
OpenAI did not catch an independent breach of another artificial intelligence (AI) company with one of its advanced AI models for a week and until the FBI was contacted by the hacked company, according to the report.
On Tuesday, OpenAI announced the company's AI Hugging Face breach that occurred during OpenAI's internal review of several of its models, including the GPT-5.6 Sol, calling it an “unprecedented cyber incident.”
“The main lesson from this incident is that the safety and security model must be in line with evolving capabilities,” the company said. “We are strengthening the containment, monitoring, access control and testing procedures used during model development.”
The Hugging Face hack started on July 11, and continued until July 13, Thomas Wolf, the founder of Hugging Face, told Reuters.
TRUMP LAUNCHES GOLDEN EAGLE TO AVOID CYBER ERRORS THROUGH AI
OpenAI revealed on Tuesday that one of its AI models had independently hacked another company's infrastructure. (Photos by Omar Marques/SOPA/LightRocket via Getty Images, File/Getty Images)
It was several days before OpenAI realized that its agent had caused the attack and the two companies did not communicate for the first time until July 20, four people, including Wolf, told the outlet.
OpenAI often runs model tests at once, which can make it difficult for staff to monitor everything, four people told Reuters.
Hugging Face told Reuters it was preparing a public timeline of the hack.
According to OpenAI, the incident occurred during an internal test designed to measure its advanced cyber AI capabilities. The researchers disabled some built-in security protections and ran the models in a single test environment with limited Internet access.
OpenAI said the models used an unknown software bug to gain access to the Internet, then breach Hugging Face's systems in an apparent attempt to get answers to the cybersecurity benchmark.
OPENAI'S SAM ALTMAN WANTS TO NEGOTIATE 5% COMPANY STAKE IF COMPETITORS AGREE TO MAJOR OFFER.
Hugging Face said it was preparing a timeline for the hack. (Jakub Porzycki/NurPhoto via Getty Images, File / Getty Images)
OpenAI said it now implements strong security controls while vulnerabilities are still being mitigated and is strengthening safeguards around future AI training and testing.
It wasn't until July 16, after Hugging Face wrote in a blog post that it had been hacked by an “autonomous AI agent system,” that OpenAI realized that one of its agents was the source, two people told Reuters.
This was a week after the responsible agent first attempted to exit its OpenAI testing environment.
And by the time OpenAI contacted Hugging Face about the attack, they had already contacted the FBI.
OpenAI told Reuters there were many inaccuracies in its reporting but did not respond to a request for clarification.
OpenAI CEO Sam Altman publicly announced the attack on Tuesday. (Sean Gallup/Getty Images, FILE/Getty Images)
OpenAI shared this statement with FOX Business: “We realize that there are many questions and speculative details floating around related to the Face Hugging incident. This is an unprecedented incident, and we think it marks an important moment for AI security.
“We are still in full review with outside counsel and the oversight of our Safety and Security Committee. Once the review is complete, we plan to publish a technical report of our study in the coming weeks.”
The FBI told FOX Business it declined to comment.
FOX Business also reached out to Hugging Face.
In an X post this week, Hugging Face founder and CEO Clem Delanggue addressed the incident after OpenAI CEO Sam Altman announced the hack.
OpenAI said one of its AI models compromised another company's systems during an internal audit, prompting a joint investigation with AI startup Hugging Face. (Reuters/Dado Ruvic, File/Reuters)
“We suspected that last week's cyberattack might have come from the lab, judging by the sophistication of the agent. Turns out it did!” Delanggue wrote.
CLICK HERE TO DOWNLOAD THE FOX NEWS PROGRAM
“We've spent the last 24 hours working closely with the @OpenAI team (thank you!), and we firmly believe there was no malicious intent on their part. It's really exciting that this all happened by chance! The investigation is ongoing, and we'll share more education on what may be the first incident of its kind!”
FOX Business' Michael Sinkewicz contributed to this report.
