This congresswoman wants to save the US from the dark side of AI

Many people feel the U.S. government has not done enough to anticipate and invest in the rise of artificial intelligence. Rep. Robin Kelly, a Democrat from Illinois, is trying to fix that.

In recent weeks, Kelly and Rep. Will Hurd, a Republican, have announced a multipronged AI initiative involving hearings, white papers and, eventually, legislation to help the U.S. cultivate a comprehensive plan for the future of robots, the Internet of Things, facial recognition technology and more. Without this level of preparation, Kelly says, the U.S. is leaving its workers, as well as its geopolitical status, at risk.

Sign up for Protocol newsletters

"I want to make sure that my folks are not left behind," she told Protocol in a recent interview. "And then on a larger level, [make] sure that AI is a positive tool, not something negative that feeds into biases and the privacy issues and those kinds of things, and that it helps us stay prepared for whatever national security issues might arise."

Protocol spoke with Kelly about the worst-case scenario she envisions for the future of AI and what she thinks Congress can do about it.

This interview has been edited and condensed for clarity.

What do you think the Trump administration has gotten wrong in its approach to AI?

I would just say that the reason that we did this is because we feel that we need a comprehensive strategy. AI is not only coming, it's here, and we need to invest more in it, invest more in our folks so people are prepared and people have skills and people know how to work with AI or have jobs in AI.

I think it's more that we still need the administration to look at investing more in artificial intelligence and preparing the country. Even though it's already here, I still feel that people are afraid of it or don't want to think about it, or they just think, "Oh, AI's going to take my job away." Some jobs it will take away, but we need to prepare people for the jobs that AI will create.

Representing the Chicago area, even though we still do have a lot of manufacturing, it's more-advanced manufacturing. We have to train people, that's the point, to get the workforce ready. People have lost their jobs because of more-advanced manufacturing; they need to be more familiar with technology.

Your white papers have focused on AI in the workforce and national security so far. Why are these the areas you decided to focus on? And what topics will come next?

Will [Hurd], last Congress, was the chair of the [House Oversight] IT subcommittee, and I was the ranking member, so we had hearings on artificial intelligence, and these were some of the seeds that came out: workforce, national security, privacy and bias.

The idea is that these white papers will eventually be accompanied by some legislation that gets at the proposals in them?

That's what we're hoping, that this is just setting the table or a foundation for what's to come.

What is the worst-case scenario for you when you think about the future of AI in the U.S.? What's the future you're trying to avoid?

I want to make sure that my folks are not left behind, that they are qualified for the jobs that are coming. That's on a smaller level. And then on a larger level, making sure that AI is a positive tool, not something negative that feeds into biases and the privacy issues and those kinds of things, and that it helps us stay prepared for whatever national security issues might arise.

Your white paper about national security and AI focused a fair amount on China and how we need to improve our AI strategy in the U.S. in order to stay ahead of rivals like China and Russia. Some fear the recent efforts against Chinese tech seems to be far more about nationalism than actual U.S. national security concerns. Do you see any validity in those concerns — that we're just at the point of engaging in blind anti-China policy actions to both countries' detriment?

I do think we need to be careful. I know my farmers [in rural Illinois] want China as a trading partner. China is involved in the United States, so again, I hate to just say, "It's Chinese, it's no good," just to be that blunt about it, but also, China has given us reason to take an extra-hard look or investigate. But also there are other parts of our society that work well with China or have suffered because we've added extra tariffs and things like that. There's always a balance.

There are many questions around whether U.S. tech companies should partner with the military and with the police at all, as critics say their technology can be used to supercharge violence, harassment and discrimination both here and abroad. What is your response to those concerns?

High tech is how [the police] solve some of the crimes they solve. Just to ban [tech companies from working with] them completely, that would be hard to do.

I've been for gun safety reform since I got to Congress; I ran on that over seven years ago, and there's not just one way to look at it — it's not going to be just one thing that changes what we're dealing with. I'm looking at it saying, yes we need background checks, we need better police community relations, we need to give police the tools so they could more accurately see where shots have come from. We need to approach this in all different ways.

There were a few, narrow provisions related to facial recognition tech in the recent police reform package from Democrats. But do you think Congress is going to be able to get together comprehensive facial recognition legislation anytime soon? What's the latest you've heard about it?

We have to get it together at some point. That's part of this whole Black Lives Matter: police accountability conversation. We have to get that together so that African Americans, Black and brown people, are not singled out or blamed for crimes. That has to be a part of reform and what we look at, and making sure what information is let in is as unbiased as possible. Again, that's the importance of having a diverse workforce. Because then what is fed in is more diverse and more wide-ranging.

Are you optimistic that Congress could get together facial recognition legislation this year? Or what's the timeline?

This year's hard to say because of who's sitting in what seats, but if something changes, which I hope it does, then it will be a more positive sign.

For a while now, you've been raising concerns that the Small Business Association programs prevent equity investment in women- and minority-owned businesses. Can you talk me through the gap you're seeing in how the SBA programs are constructed and how you want to fix that?

It's really a simple rule, that if someone loans you money or makes an investment into your company and they say that they want a percentage of it — the "shark tank rule" — that people would be allowed to own a piece of the company without disqualifying the primary owner from being ineligible for SBA assistance.

That impacts women more because we need help with owning our businesses, getting our businesses off the ground, and to say that it has to be free and clear of anybody else, that affects our ability to grow our business. [The SBA currently requires businesses to be 51% unconditionally owned in order to qualify for particular aid programs. That rule can force businesses to choose between equity investments and government help.] We're still beating the drum.

Are there any tech- or cyber-related provisions that you most want to see in the next COVID-19 relief package?

There are big pushes for telehealth and telemedicine. We want to see that continue. More of it was done and then some of the rules and regs were waived. We have a bill that studies the effect of what has happened over this period of time, but also there are other bills that are more aggressive, like, "Let's just waive the rules and regulations." Telehealth done properly can be the great equalizer as far as racial disparities that we have in this country.

When we spoke to a group of Black doctors across the country, they talked about how patients, because of telemedicine, were able to keep their appointments and didn't have to worry about jumping on a bus, they didn't have to worry about child care. Then in my rural areas, telemedicine has definitely been used, particularly around behavioral health or counseling or therapy. We need to do whatever we can to enhance — again, if done right — the whole aspect of telemedicine.

What are the specific provisions you're pushing for on telehealth?

The payment being the same whether it's a face-to-face visit or a telemedicine visit — that's one thing that's certainly helped.

Tech companies are notoriously white and male, and they're not making much progress in changing that, despite their outward pledges to diversify. What role do you think the government should play in ensuring there's less discrimination and more diversity in tech?

We've had legislation where the government helps as far as scholarships or loan forgiveness for people [who] have gone into the STEM fields and tech fields. But also, the other thing is, a major role we can play, which we're trying to play through the infrastructure bill: We have to make sure there's broadband all over. In my district, some people have a very hard time getting online. There's some fundamental things we have to do so people even realize they can go into the tech field.

What is the future of privacy legislation in the U.S.?

I've battled myself with privacy because I really believe in one's privacy, but then it's tough when you're trying to solve a crime, or if you think the nation's security is at risk or question. I'm sure we'll delve back into it. Right now, COVID has taken over. We have to take care of the matter at hand. But I don't think that's something that's going to completely disappear off our radar. There'll be something that comes up anyways that brings it back.