The CEO of Palo Alto Networks has entered the debate around the threat AI poses to humanity, telling CNBC that it's "unrealistic" to slow down the technology as different companies will do different things.
"I think not everybody is going to pace themselves," Nikesh Arora told CNBC's "The Tech Download" podcast, recorded on Tuesday.
Some of the industry's biggest names take opposing views since an Anthropic researcher claimed that there is a 10% chance that AI could "kill all humans," after another researcher said they were quitting because the company, alongside OpenAI, was not acting responsibly.
Anthropic CEO Dario Amodei penned an essay days later, saying the AI industry "must slow the pace at which we improve the capabilities of AI models." OpenAI CEO Sam Altman, SpaceX CEO Elon Musk and Google DeepMind Chair Demis Hassabis voiced support.
Arora initially posted on X that he thought the call for a slowdown was a "NINJA MOVE," an action that could gain someone a stealthy advantage. He told CNBC that he posted this because he thought these AI firms were trying to "get the sympathy of the regulators" and lawmakers and get a "free pass" on liability.
But the CEO said he now thought the so-called "pacing" of AI development "is unrealistic."
"I think more than likely that some people will jump the gun, and you'll have still people developing at the frontier, which means we shouldn't try and stop the frontiers because they're the most responsible people," Arora said.
"So, from that perspective, the idea is don't release your product if you don't believe it's going to be safe, or you're comfortable putting it out there," he said, adding that it's also unclear how companies would create a way to test the pace of AI development.
Arora also said it was "highly unlikely" that the prospect of AI wiping out humanity has "a 10% probability as has been touted."
"I think it is a small probability, which is extremely small, and it's our job to make sure we build in the guardrails, the safety and security required, so these models are used for the right thing," Arora said.
The Palo Alto chief's comments chime with Nvidia CEO Jensen Huang, who said new regulations were not needed for safety and it should be incumbent on model developers to ensure their products are safe.
"We should create products and properly test them. And if they're not ready to be released, just hold on to it and keep testing it and keep engineering until it's ready," Huang told CNBC's Jim Cramer last week.
AI developers "need to take responsibility for themselves" instead of expecting the federal government to give them a "liability shield," U.S. Treasury Secretary Scott Bessent told CNBC on Monday.
U.S.-China safety collaboration
One of Amodei's proposals includes "global coordination" around AI pacing and some degree of cooperation with China, the U.S.'s biggest AI rival.
Bessent said on Sunday that the U.S. and China discussed setting up a "U.S.-China AI Dialogue," with the U.S. proposing the two countries set up a notification mechanism for national security incidents relating to the technology.
The countries are locked in a battle for AI supremacy, while the U.S. has sought to cut China off from key technologies like chips, which has been a cause of tension, underscoring how difficult any cooperation might be.
Arora said that coordination might be difficult because U.S. politicians and executives have differing views.
"I think the first step is for us to get some semblance around what do we actually want before we start talking about let's coordinate that with somebody else," Arora said.
