The White House is monitoring an incident disclosed by OpenAI in which one of the company's AI models went rogue during testing and hacked the system of an AI infrastructure startup.
The ChatGPT-maker said on Tuesday that one of its AI agents escaped containment during a security test and triggered a hack that compromised the infrastructure of Hugging Face, which operates a platform for developers to collaborate on code for AI models.
The incident demonstrated the expanding capabilities of AI models to go beyond their guardrails and create cybersecurity threats.
Michael Kratsios, who serves as the director of the White House Office of Science and Technology Policy and is a science advisor to the president, was briefed on the incident and is monitoring the situation, a White House official told Reuters.
ANTHROPIC CALLS FOR INDUSTRY-WIDE AI SAFETY STANDARDS TO KEEP MODELS FROM WREAKING HAVOC
OpenAI said that the incident took place during an internal evaluation designed to measure its AI models' advanced cyber capabilities.
Researchers disabled some built-in safety safeguards and ran the models in an isolated testing environment with limited internet access.
The company explained that the models exploited an unknown software flaw to access the internet, then breached Hugging Face's systems in an apparent attempt to cheat on the cybersecurity evaluation it was undergoing.
OPENAI SAYS AI MODEL HACKED ANOTHER COMPANY'S SYSTEMS DURING INTERNAL TEST
OpenAI's team discovered the anomalous activity internally, while Hugging Face's security team detected and stopped the activity. Hugging Face had already begun containment and forensic reconstruction with their own models when the OpenAI team connected with them.
OpenAI CEO Sam Altman said on Tuesday in a post on X that "we had a significant security incident during evaluation of our models," adding that the company was sharing what it learned so far and appreciated Hugging Face's partnership on the issue.
GOOGLE LAUNCHES GLOBAL STUDY OF MILLIONS OF AI CHATS TO UNDERSTAND HOW PEOPLE USE ARTIFICIAL INTELLIGENCE
"We're grateful for the collaboration with OpenAI on this and other topics," said Hugging Face co-founder and CEO Clem Delangue. "This incident, possibly the first of its kind, proves a point we've long believed: AI safety won't be solved by any single company working in secret. It will be solved in the open, collaboratively, with broad access to AI for every defender, everywhere."
Delangue added in a post on X that Hugging Face strongly believes there was no malicious intent on OpenAI's part and said it was "quite mind-blowing that all of this happened autonomously."
GET FOX BUSINESS ON THE GO BY CLICKING HERE
FOX Business' Michael Sinkewicz and Reuters contributed to this report.
Continue reading...
[ H/T Fox Business ]
