EXCLUSIVE: Former Google CEO Eric Schmidt Butts Heads With Former FTC CTO Over AI Regulation

In a heated discussion with leading AI scholars Tuesday evening, Eric Schmidt, the American businessman who held the top post at Google from 2001 to 2011, argued that AI systems can develop unexpected behaviors that limit the extent to which companies like Google can implement preemptive safety and governance mechanisms into their products.

A central “problem” with regulating frontier AI models, is that, sometimes, “a new feature emerges in these systems that is not tested, testable,” Schmidt said onstage at the annual Isaac Asimov Memorial Debate, moderated by physicist Neil deGrasse Tyson in Manhattan. 

“We can stop [the emergence of new features or behaviors], and therefore stop all progress, by law, by banning larger models,” he said, “but as long as you have this new emergent power, you have deep reasoning, deep capabilities, and they will make mistakes. You have to be tolerant.”

Schmidt—who also served as chairman of Google’s parent Alphabet from 2015 and 2017 and advised Alphabet for three years after that—supported the 2014 acquisition of DeepMind, the unit that now houses Google’s most cutting-edge AI research. 

Schmidt said that AI developers, like Google and others “should be held accountable” if they’re found in violation of the law, but emphasized that AI developers frequently have to ship AI products and retroactively correct bad behaviors they didn’t predict as the models evolve. 

“I went through this when I was at Google, in earlier versions of [Google’s AI] technology, where the system would actually do something that was wrong, and we fixed it. And we fixed it as fast as we could, because we had to, because it was the right thing to do,” he said. 

Google did not respond with an on-the-record statement by press time. 

Schmidt was challenged by Latanya Sweeney, a professor of government and technology at Harvard and the former chief technology officer at the Federal Trade Commission, who cast doubt on the suggestion that AI leaders would happily comply with regulations. She said that leading tech companies have proven time and again to ignore key regulations or attempt to bend the law to their commercial interests. 

Sweeney is not off base. Just last year, U.S. federal judges found Google guilty of operating not one but two illegal monopolies—one in adtech and one in online search. The decisions came six years after Meta, then Facebook, was forced to cough up $5 billion to the FTC and modify its business practices after it was found to have mishandled reams of user data in the now infamous Cambridge Analytica scandal.

“Technology just ignores [laws] and rewrites them. That’s true in social media. It’s true in AI as well,” Sweeney said to Schmidt. “We already have laws [that] address issues of bias, consumer protection, and so forth. None of those are enforced online.”

Sweeney suggested that mitigating the risks associated with AI, including algorithmic harms and biases encoded into AI systems from their training data, requires more foundational changes rather than ex-post-facto fixes. 

“There are questions about existential harms in the future, but there are a lot of harms happening right now. And it doesn’t have to be that way….It depends on who this AI is servicing, and in particular, the design of the technology—the decisions made in that design is really determining what our values will be,” she said.

Schmidt pushed back on the implication that all harms could be preempted with better design and training, arguing that leading AI programs are not simple machinery but complex, non-linear systems that often develop unforeseen capabilities. But he agreed with Sweeney’s assertion that Silicon Valley leaders have at times “rushed” products to market, adding, “They’ve found all sorts of problems, and then they’re busy correcting them. I think that’s the cycle, and it’s very hard.”

Luckily, he said, AI developers have evaluation cards and safety testing teams in place to mitigate as much risk as possible in advance. 

But these measures are insufficient for protecting humanity, according to another panelist, Nate Soares. Soares is president of the Machine Intelligence Research Institute in Berkeley, California and co-author of If Anyone Builds It, Everyone Dies, a 2025 book on the existential risks posed by AI. 

In leading AI labs, the primary focus areas for safety and governance today are interpretability research, or “trying to figure out what’s going on inside the AIs’ heads,” and model evaluation cards, “which are trying to figure out how dangerous the AIs are,” as Soares explained.

He likened these efforts to a comically inadequate attempt to curtail nuclear disasters. “If someone was making a nuclear power plant in your hometown, and you went to them and you said, ‘Hey, I hear that this uranium stuff can have lots of energy benefits, but also can melt down when things go badly. What have you guys got that makes you think you’re going to get the benefits and not the pitfalls?’ If the engineers say—‘Oh yeah, we’ve got two crack teams working on this; the first team is trying to figure out what the heck is going on inside, and the second team is trying to measure whether it’s currently exploding’—that’s not a good sign.” 

Soares expressed urgent concern about the future of AI for the same reason Schmidt found the technology to be too difficult to govern: because frontier AI models tend to act in ways their makers didn’t explicitly design or even anticipate.

Schmidt argued that the benefits of AI will ultimately outweigh the risks, disagreeing with arguments made by Sweeney, Soares, and other academics onstage. 

“The companies that are doing this work are well aware of the dangers, and I know this because I work with them on this, and we spend an awful lot of time talking about them,” Schmidt said. “We can talk about some of them, but that is not to take away from the enormous benefit of these technologies.”

Schmidt, deGrasse, Sweeney, and Soares were joined onstage by Kate Crawford, a professor of AI at the University of Southern California, Chris Callison-Burch, a University of Pennsylvania computer science professor, and Cindy Rush, a statistician who teaches at Columbia University.