JUANA SUMMERS, HOST:
Picture a cartoon of a fox dressed as a security guard, standing in front of a henhouse. Consider that image alongside the news that this week, President Trump invited a group of top AI executives to the White House to announce a voluntary AI accord that would involve, quote, "tremendous self-policing" and would not be legally binding but morally binding. Many AI leaders are concerned that the technology may be developing too quickly to address its safety risks, such as AI agents going rogue and hacking other companies or entities.
Well, that cartoon - it appears with The New York Times op-ed by Lina Khan, chair of the Federal Trade Commission in the Biden administration. The title, "There Are Plenty Of Laws On The Books To Check The AI Giants. Use Them." And Lina Khan is with us now. Hi there.
LINA KHAN: Hello.
SUMMERS: Lina, I mean, there's so much we can talk about, but I want to start with this accord to self-police AI development. Can this actually do anything useful, or is this really just more like the fox guarding the henhouse?
KHAN: It's exactly like the fox guarding the henhouse. I mean, this idea that you have these voluntary commitments that everybody has signed off on is just laughably toothless, especially given the very serious risks and dangers that these companies themselves are noting we could face.
SUMMERS: Now, there is some action from federal regulators. Yesterday, the Federal Trade Commission, which, of course, is your former agency, confirmed to the Associated Press that it had opened an investigation into companies like OpenAI and Anthropic over the risks their products could pose to consumers. What do you make of all that?
KHAN: Well, look, I think what we've seen under this administration is extraordinary levels of backsliding on law enforcement when it comes to holding accountable wealthy, powerful interests. And so the idea that you're going to have an agency be investigating - seriously - the very CEOs that are having private lunches with the president, especially what we've seen from this White House in terms of giving pardons to corporate lawbreakers and CEOs, even when they've committed very serious violations of the law, just seems to be somewhat incredulous.
SUMMERS: Now, you write that there are plenty of regulatory laws that apply to AI companies - things like product liability law, tort law, consumer protection laws, state laws, individual liability of AI executives. What harms of AI do you hope to see addressed by enforcement of these laws?
KHAN: I mean, I think the big-picture point here is that a lot of the conversation right now around these companies and some of the harmful practices that are resulting from their technologies, seems to assume that this all is happening in some kind of legal vacuum. And that's just not true. Just because you have a new technology come to the market doesn't mean that all existing laws are suspended or...
SUMMERS: Right.
KHAN: ...There's some kind of exemption. And so, you know, laws that prohibit unfair, deceptive practices, laws that make it illegal to hack into other websites, laws that make it illegal to sell defective products - all of those very much apply to these technologies. And so as we are seeing more and more incidents of, you know, some of these systems engaging in serious hacks - be it of other companies, of government servers - that's all something that absolutely needs to be investigated under existing laws, some of which have criminal liability.
SUMMERS: Yeah. I mean, we all know the legal system can often move, super, super slowly. And we also know that the tech companies - they have enormous legal resources. We've seen this play out often with lawsuits against social media companies. Are you worried that this might just take a ton of time?
KHAN: It could. Although some of what we've seen from, for example, the Florida AG, where they're going to court to get an injunction against some of these very practices, could set us up to get answers more quickly.
But you're absolutely right that, overall, what we need is, you know, a broader regulatory regime. And so one additional thing I note in the piece is that we actually have a long history of regulating and governing new technologies...
SUMMERS: Yeah.
KHAN: ...That are powerful and that are risky, from railroads, to pharmaceuticals, to banking, to nuclear power. And so we should really use the lessons from those regulatory regimes as we think about what to do here.
SUMMERS: I'll also note that you say that Congress will need to pass further laws. Like, what? Give me one example.
KHAN: So, for example, there's a long history in banking law, for example, of this thing called supervision, where you have the government, you know, deeply involved in helping monitor risk inside banks to make sure that they're not going to end up, you know, creating huge financial crashes. Similarly, as we're thinking about pharmaceuticals and drugs, there's all sorts of mandatory testing that needs to happen before a pharma company...
SUMMERS: Yeah.
KHAN: ...Can hit, you know, the market. And so those are some of the types of things that we need to be looking at here.
SUMMERS: We've got about 30 seconds left here. I'll just note we live in a very historically dysfunctional Congress. We're living with that. We also have a president who says we don't need any more guardrails for AI. We need to win the AI development race with China. Any fear that your calls for more regulation and more enforcement will go ignored, at least for the next two year while President Trump is still in office?
KHAN: Well, look, I think this administration has shown that they have no real interest in enforcing the law against powerful companies. And so I do think that a lot of the action will need to come from Congress, be it at the level of oversight and investigations. There are all sorts of questions we need answers to.
SUMMERS: Yeah.
KHAN: And this is also why state AGs have an incredibly important role to play.
SUMMERS: Lina Khan, former chair of the FTC, now at Columbia Law School and chair of the New York City Economic Development Corporation. Thanks.
KHAN: Thanks. Transcript provided by NPR, Copyright NPR.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.
