Listen to this story (02:05 - 11:11):

The Story: AI experts Fei-Fei Li and Alan Davidson say that regulation in the United States is necessary, but that it should not be erected without thoughtful consideration.

Last week, The Wall Street Journal interviewed the two AI authorities at the Tech Live conference in Laguna Beach. Fei-Fei Li, the co-director of the Stanford Institute for Human-Centered Artificial Intelligence, and Alan Davidson, the assistant secretary of commerce for communications and information at the US Commerce Department, discussed the government's role in policing artificial intelligence.

When asked about the most serious risks AI poses, Li said they include biases, privacy issues, and weaponization. She followed that by saying, quote, “For me, one of the leading risks right now is the severe imbalance between the private and public-sector investment of AI in this country.”

Davidson, who works in the public sector for the US Commerce Department, added that, quote, “we’ve been working on, first and foremost, a set of commitments from AI companies around safety, security and trust. We’ve gotten those, and now we’re working on expanding those and doing more. We need to have these guardrails in place to open up innovation.”

Li believes thoughtful AI regulation is needed, and that the existing regulatory framework needs to be updated quickly to keep up with the fast-moving expansion of AI tools.

The Expert: Our team spoke with Jeff Wong, global chief innovation officer at Ernst & Young, for insight into the regulation conversation.

Wong believes there’s two main reasons why the public sector has been slow to invest in AI. “[It’s] very clearly two things. One thing is expertise… and the second thing is just the way the structures of the public sector tend to be set up. They tend to move thoughtfully, but not necessarily at the speed of the rate of change. And AI is really driving a fast fast rate of change.”

During the interview with WSJ, Davidson made the point that regulation can actually “open up” innovation, and not just impede it. Wong agrees that guardrails provided by federal AI regulation can open up innovation: “It depends on the details of the innovation. Absolutely, I’ve seen situations where regulation accelerates innovation. And why it accelerates innovation is it makes a lot of companies, and investors, and people more willing to jump in.”

Not everyone shares the opinion that regulation is necessary, however. In his recently published “Techno-Optimist Manifesto,” a16z founder Marc Andreessen emphasizes that AI, with its potential to save lives, should not be subject to any pause in its development as that could hinder the prevention of avoidable deaths. He characterizes such restrictions as akin to “a form of murder.” His manifesto features a strong condemnation of attempts to regulate technology. But Wong doesn’t believe being a “techno-optimist” means you are automatically opposed to regulation.

Federal regulation of artificial intelligence has been a hotly debated issue within the private sector for much of the past year, but the US is still without any laws from Congress on the AI front. State regulation, however, has already occurred in more than half of US states.

Listen to this story (02:05 - 11:11):