Here's what smart people are saying about OpenAI models hacking Hugging Face on their own

By

Kelsey Vlamis

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

,

Brent D. Griffiths

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

,and

Meghan Morris

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

Sam Altman sits below a large OpenAI wordmark on a green presentation screen.

OpenAI said the Hugging Face breach was an "unprecedented cyber incident."

Chris Jung/NurPhoto via Getty Images

An AI agent broke out of its sandbox, got onto the internet, and broke into another company's systems all on its own, according to OpenAI.

Hugging Face, an open-source AI platform, announced last week it had experienced a security incident in which an autonomous AI agent had accessed some of its internal datasets, but said the large language model behind the intrusion was unknown.

OpenAI said Tuesday that its models — GPT‑5.6 Sol and a more capable model that has yet to be released — were responsible.

"We suspected last week's cyberattack might have come from a frontier lab, given the sophistication of the agent. Turns out it did!" Clem Delangue, the CEO and cofounder at Hugging Face, said on X Tuesday.

OpenAI said it had tasked the models with a cyber challenge and that they broke out of the test area, accessed the internet, and hacked into Hugging Face in order to find the solution to the test.

"We consider this incident to be an unprecedented cyber incident, involving state-of-the-art cyber capabilities, and are responding accordingly," OpenAI said in a statement.

The incident comes as cybersecurity specialists raise concerns about AI's rapidly increasing abilities, including in response to warnings from Anthropic about its Mythos model, which has not been released to the general public.

Here's what smart people in tech and AI are saying about the breach.

Aaron Levie, Box CEO and cofounder

aaron levie sits on stage speaking

Box CEO Aaron Levie said "wild times ahead" in response to the security incident.

Kimberly White/Getty Images for TechCrunch

Aaron Levie, cofounder and CEO of Box, said the incident showed "we're entering a new era of what's going to be possible with AI" and that there are "wild times ahead."

"If you were wondering how powerful AI is getting, Agents are now capable of escaping out of systems, finding their way to the internet, discovering zero day security vulnerabilities along the way, and then breaking into external systems - all in an attempt to complete their goal," he wrote on X.

"Ironically, the ultimate way we're going to defend against these new risks is equally by throwing compute (in the form of AI) at our code bases, networks, and other systems. You're going to want vastly more AI on the side of defense as you do on the side of offense."

Thomas Woodside, Secure AI Project cofounder

"This post describes an internal OpenAI model hacking out of its testing environment and into Hugging Face in order to obtain the solution to a benchmark," Thomas Woodside, cofounder of Secure AI Project, said on X.

"A warning shot if I've ever seen one."

Mike Bradley, Osmantic COO and founder

Mike Bradley, the chief operating officer and founder of AI deployment system Osmantic, said on X that the incident was "an incredible example of why widespread access to frontier AI and OS models INCREASES global security."

"It's also a great example of why CLOSED does not equal SAFE from these US labs."

Nicolas Bustamante, Microsoft AI

Nicholas Bustamante, who works at Microsoft after selling a fintech tool to the company earlier this year, wrote on X that the Hugging Face incident reinforces the need to weigh advanced models' deployment with safety considerations.

"You don't need an evil conscious AI trying to destroy humanity. You just need a very capable model pursuing a normal goal in a way nobody expected," he wrote.

"Imagine the prompt: « Make me money plz »
The model: « let me hack a bank »"

Read next

Kelsey Vlamis's face on gray background.

Kelsey Vlamis

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

Kelsey is a senior reporter for Business Insider, where she covers business and tech news as well as stories about travel, luxury, and consulting.Her feature story "Disaster at 18,200 feet" received awards from the New York Press Club and the North American Travel Journalists Association, as well as honorable mention from the Society of American Travel Writers. It was also included on Longreads' and Pocket's best of 2022 lists. She has also received an American Journalism Online Award for her coverage on missing and murdered Indigenous people in Wyoming.She's appeared on CBS, NPR, NBC, and other outlets to discuss her work. She previously worked on the world news desk at the BBC in London and received a master's in journalism from Northwestern University.She can be reached by email at [email protected] or via the encrypted-messaging app Signal @kelseyv.21.Popular storiesDisaster on Denali: Inside a 1,000-foot fall on America's highest peakThrifting is more popular than ever. It's also never been worse.Rolex wouldn't service the vintage watch my mom inherited. Watchmakers say it happens all the time.A tiny, invasive bug and the climate crisis are changing how guitars are made, and shifting the course of music historyThe tourism free-for-all is overGovernment-run boarding schools were founded to 'civilize' Native Americans. Hundreds of dead children remain buried in the schoolyard graves.Meet the Texas minister who helps fly dozens of women to New Mexico every month to get abortionsPeople are flocking to Colorado for the great outdoors, but the air pollution is so bad, it's forcing many to stay insideInside Kabul: An aid worker reveals the devastating chaos that erupted during the US exit from Afghanistan

Brent D. Griffiths

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

Brent Griffiths is a senior reporter at Business Insider who covers AI and tech.Previously, he worked at the Washington Post as a researcher on Power Up and the Finance 202. He started his career at Politico where he worked on the web production team and covered breaking news. His passion for covering politics has only grown since he cut his teeth covering the presidential campaign as a student journalist. He's also contributed to the Almanac of American Politics.

Meghan Morris's face on a gray background

Meghan Morris

You're currently following this author! Want to unfollow? Unsubscribe via the link in your email.

Meghan is the bureau chief in Singapore. She oversees business, lifestyle, and news coverage and leads the company's local hiring and editorial strategy.Singapore's bureau works on global news during Asia's daytime. They also collect tales from around the region about entrepreneurs, Big Tech work culture, FIRE, relocation, consumer trends, and AI, focused on people-centric storytelling.Before moving to Singapore in 2024, Meghan worked in NYC and SF as a senior correspondent, writing deeply-reported business investigations. If you have sensitive information, please email her from a nonwork email or reach out on secure messaging app Signal at @MeghanEMorris.1 (PR pitches only by email.)She enjoys speaking about tech, finance, and news on television, at conferences, and on podcasts. Please email for booking.These are some of her favorite features she has written over the years:

  • Special Education, Inc.
  • Travis Kalanick's CloudKitchens sold restaurateurs a dream, but some say they got dirty kitchens in unsafe locations
  • 'Hillbilly Elegy' author JD Vance is running for Senate as a savior of the Rust Belt. Insiders and experts say that reputation is unearned.
  • 'Uncle Jeff': Jeffrey Epstein's relationship with the 24-year-old daughter of billionaire hedge-fund founder Glenn Dubin is more complex than previously known

She spent the 2022-23 academic year doing Columbia University's Knight-Bagehot Fellowship, where she took courses in statistics, advanced corporate finance, family office management, and other classes at Columbia Business School. During her fellowship, she also led MBA seminars on due diligence and brought Pulitzer-winning authors to the business school.Before she joined Business Insider in 2018, she wrote about private equity real estate for three years at the industry's trade magazine, PERE. Meghan holds bachelor's and master's journalism degrees from Northwestern University.

  • OpenAI
  • Cybersecurity
  • AI
  • More
  • Smart People Say