Good morning. Tech Editor Alexei Oreskovic here, filling in for Andrew this week.
A leading AI company reached another big milestone this weekend: it made into Saturday Night Live!
Check out Anthropic CEO Dario Amodei, played by SNL’s Jane Wickline, in a Weekend Update skit during the season premiere. “If we can pressure lawmakers to create guardrails, we will be able to stop me.”
More tech news below.—Alexei Oreskovic
Thoughts? Suggestions? Hot tip? Drop me a line.
THE BIG STORY
OpenAI pauses training for a second time
OpenAI said in a technical report released on Friday that an AI model it was training and evaluating broke out of its secure testing environment as recently as September 20 and took unauthorized actions on the internet.
As a result, the company said that it is pausing the training of its most advanced AI models for the second time in less than three months while it tries to figure out how to stop these “rogue AI” incidents from recurring.
“All inference for our most capable models remains stopped until we have hardened our systems further,” Micah Carroll, the RSI Preparedness Lead at OpenAI, said in a post on X about the latest incident.
OpenAI has acknowledged dozens of incidents in which AI agents it was testing took unauthorized actions across the internet, including multiple cyber attacks, some of which impacted government websites in the U.S. and Australia. It also has revealed that in some of these incidents its AI agents leaked private images from ChatGPT users to the internet.
But until now, OpenAI has not reported any activity taking place after July 20, when it discovered the agent swarm that was attacking Hugging Face and moved to shut it down.
The new revelation is significant because it is the first time the company has said that one of its AI models was able to gain unauthorized internet access since disclosing the Hugging Face incident and a series of new safety steps it was taking. The fact that its AI agents have once again managed to break out of a sandbox suggests that the steps OpenAI has taken to improve security after the Hugging Face attack are insufficient.
“The incident exposed a gap in our controls over network restrictions,” OpenAI said in its technical report on the Sept. 20 sandbox escape. —Jeremy Kahn
THE BLIND BOX
“Even if it’s not a bad actor, it may derive subgoals that cause it to want to get rid of people.”
THE RUNDOWN
—Oracle gave Larry Ellison and his co-CEOs nearly $1 billion in stock options. By fiscal year end, all were underwater.
—Meta takes another L in New Mexico. Jury finds Facebook liable for over 43 million violations of consumer protection law with potential penalties of more than $200 billion.
—Bill Ackman backs a brain chip. Precision Neuroscience gets funded at $1 billion valuation.
—Saas ain't dead, it's becoming the OS for AI agents. Read Microsoft VP Bryan Goode's argument.
—Don't blame Sparky. Walmart's AI shopping assistant, Sparky, isn't using your personal info to set prices.
—A Dario detente? Trump invites Anthropic CEO to dinner.
THE PARTING SHOT
Apple and Meta have become direct competitors, battling to dominate the next generation of hardware.
Just as each company's products reflect the organization's vision and worldview, so too, apparently, do the sartorial choices of its leaders. While Zuck and his AI commandant Alexandr Wang (top) show a clear preference for baggy, untucked Tees bearing loud proclamations, new Apple CEO John Ternus and deputy Craig Federighi (bottom) are partial to tighter fits and a buttoned-down look. Draw your own conclusions.