Slate of bills looking to regulate AI introduced at Minnesota Capitol

CHRIS FARRELL: Artificial intelligence, the Minnesota legislature is grappling with how this fast-growing artificial intelligence market should be regulated by the state. So, today, the Senate Judiciary and Public Safety Committee is hearing a package of bills from State Senator Erin Maye Quade that would put regulations on how AI can be used in everything, from health insurance decisions to chatbots used by children, to dynamic pricing in the marketplace. DFL State Senator Erin Maye Quade from Apple Valley is on the line to talk about it and thank you for joining us.

ERIN MAYE QUADE: Thank you so much for having me.

CHRIS FARRELL: So we all know, artificial intelligence is moving incredibly fast right now, it feels like, in every field imaginable, whether it's healthcare or law or journalism. So what made you decide this was the moment for the legislature to start putting some guardrails in place?

ERIN MAYE QUADE: So, candidly, I've been working on regulating big tech and AI for my entire first term, in all four years that I've been here. I think what's changed is that there's an appetite amongst the broader legislature to do something. I think the issues that we're seeing, I've called it a five-alarm fire-- I stand by that-- I think folks are starting to see the smoke and feel the fire. And it's really time for states to step up in the absence that's been left by federal Congress. And that's happening in red, purple, and blue states, and I'm so proud to be part of that coalition.

CHRIS FARRELL: So why a five-alarm fire?

ERIN MAYE QUADE: Yeah, so when you hear from maybe AI companies and they talk about the promise of AI, and I actually think that the promise could be tremendous. I think that we could see a cure of cancer or solving things like poverty. But that's not what this technology has been rolled out to do. And, in fact, they've had devastating and deadly consequences for AI that's kind of been forced into our lives with no guardrails, no regulations.

We've seen AI psychosis, child sexual abuse material generated. We've seen children dying. We've seen our constant data surveillance put our constitutional liberties at risk.

And, increasingly, we're just interacting with computers when we should be interacting with humans, unbeknownst to us. And this is a place where the legislature can step in and say, hey, we need some guardrails, some common-sense guardrails. These bills are bipartisan, and I think we're going to have success today.

CHRIS FARRELL: So are you learning from other states? Are you talking to other states because there's legislative initiatives going on around the country?

ERIN MAYE QUADE: Yes, I'm connected to legislators of both parties across the country, from Utah to North Carolina to Tennessee to Massachusetts to New York, where we are advancing these policies. And it is a space of just real, true bipartisanship. We are on a Signal thread together. We share our bill ideas.

We share blocks that we run into or arguments that people have raised, so we can help each other get these passed. And it has just been a very, very good space for state legislators to be in, working together at a time where we see a lot of disagreement between the parties, and that's real. But this is one place where we've had incredible unity.

CHRIS FARRELL: So why is that? That was going to be my next question. We have deep political divisions on critical issues, but there is bipartisanship here?

ERIN MAYE QUADE: There really is, and I think there's two reasons for that. One, a lot of this is about harm to kids. And we tend to find a lot of bipartisan support in ways that we can protect kids from harm in general in society.

And the second is that a lot of the things that we're talking about today in Judiciary and my other bills are taking legal frameworks that already exist for things we have in the physical world and applying it to the digital world. Price fixing is already illegal. It should be illegal if two people get into a room to do it or if they submit their data to an algorithm and the algorithm does it for them.

So I think that's also part of it is that these are commonly accepted legal frameworks. And we're just going to make sure that AI and tech apply to them as well. And the other part, too, is that the harms caused by AI know no political party, and they know no region. It really is happening to everyone.

CHRIS FARRELL: Now, I know you're under time constraints, so just very briefly, let's focus on the children's access to recreational chatbots, that section of the legislation.

ERIN MAYE QUADE: Yeah, so chatbots have been-- first of all, I have a three-year-old. And AI chatbots are younger than my daughter, so this hasn't been around that long. And in just three short years that this has been around, we have seen numerous people die by suicide.

We've seen many, many instances of AI psychosis. And this is in adults and these are people that overwhelmingly have no previous mental health history. They have no substance use disorder. They aren't even particularly interested in technology above and beyond the layperson.

They ask ChatGPT an innocent question about math to help their eight-year-old son and, 21 days later, end up in a state of delusion that they've created a novel mathematical formula that could destroy the internet. This is a real story, and so that is just so dangerous. And you take that same concept and apply it to a developing mind, and we are seeing young children as young as nine being exposed to hypersexual content.

That would be illegal if a person were exposing them to it. We have seen young people engage in self-harm and disordered eating, encouraged by chatbots. This is just completely harmful and unnecessary. We don't need to have chatbots talk to kids about kinks in order to cure cancer.

CHRIS FARRELL: Well, thank you very much for your time. I really appreciate it.

ERIN MAYE QUADE: Thank you so much.

CHRIS FARRELL: That's DFL State Senator Erin Maye Quade.