98 teens got together to craft an AI bill. Here’s what happened.

0:00Hey there, it's Juana Summers. I want to share something real quick before we get to the show. It's a comment that an NPR Network supporter shared with us. Kim from Chicago says, I trust the news reported by NPR. It's critical to have a news source that is not controlled by corporate or political interests. It's true. NPR has no shareholders and no editorial interference. It's the public we serve, you. And public media requires a growing community of supporters to continue its mission of creating a more informed public. Our July fundraiser ends tomorrow. Help us end this month strong. Head to donate.npr.org.

0:36Thanks. It's Consider This where every day we go deep on one big news story. Today, can teenagers come up with guidance for how schools should use AI? Are there any senators that would like to do motion for a vote on Senate Bill 2226, the Students' First Act here today? Mr. President, I motion to close the vote. This summer, teens from across the country gathered in Boston. They were participating in a mock student senate. Recording of the A's and A's, clerk being myself, will call the roll. Senator Acho. The topic, artificial intelligence, and coming up with a bill that would guide schools on how to use it.

1:08When we think about AI in education, the conversation shouldn't be about whether students are using AI. That's Vahiva Ulani Gapero, a student senator from Hawaii. It should be about ensuring that they know how to use it properly and responsibly and with integrity. And I believe that this bill helps to provide that foundation. Consider this. AI is transforming the world young people will inherit. How are they thinking about how to deal with it? From NPR, I'm Juana Summers. This message comes from Cachava. It can be tough to stick to your wellness routine, especially when you're on the go. Cachava's new travel packs make it easier.

1:49Just one packet provides nutrition ready for wherever life takes you with protein, fiber, greens, probiotics, electrolytes, and more. Take your daily ritual with you. Go to cachava.com and use code NPR. New customers get 15 % off their first order. That's K-A-C-H-A-V-A dot com. Code NPR. This week on Here and Now Anytime, Democratic voters in Michigan sound off ahead of the big primary election on Tuesday. The average American, I don't feel like our voice is being heard. I feel like a little man at the bottom of the totem pole. Also, Haitian Americans in Ohio fear deportation after losing their temporary protected status.

2:30All that and more on Here and Now anytime. Listen on the NPR app or wherever you get your podcasts. Instead of scrolling mindlessly, engage mindfully with the NPR app. With a mix of on-demand news, stories from this station, and your favorite podcast, you can relax without shutting off your brain. Download the NPR app today.

2:58It's Consider This from NPR. Over the course of one frantic summer weekend, 98 teens representing all 50 states met in Boston to try to do something remarkable. Agree on one unified national policy to help schools navigate the chaos created by artificial intelligence. NPR's Cory Turner has their story. These teens, most of them high school juniors and seniors, came from big cities and small towns, blue states and red states. And they had at least one thing in common. When it comes to AI, they want guardrails. Without that, it's just going to go unhinged. Everyone's going to stop learning and start relying on AI.

3:36Instead of using AI as a tool, they're going to use it as a brain. Campbell McHenry goes to school in Mississippi, and he's talking here about what researchers call cognitive offloading. I asked a lot of the teens about this. And is that something that you guys have seen? That there are plenty of kids who are just looking to get through an assignment? Oh yeah. I'm friends with most of them. I've got assignments that people just go, yeah, as soon as I get home I'm just putting this whole thing into chat GPT. And that's partly why Colton Larson of Idaho and Vivian Dodd of Wyoming came to Boston.

4:11to be a part of this mock Senate to craft an AI policy that teens and schools can agree on. I do solemnly swear that I will support and defend. I do solemnly swear that I will support and defend. The weekend began at the Edward M. Kennedy Institute for the United States Senate. In a replica of the Senate chamber, the students stood at their desks, taking the oath of office. So help me God. So help me God. You may be seated. Give yourselves a round of applause. Welcome to the United States Senate. This team's Senate was handpicked by their school leaders, and the event was co-sponsored by several non-profits, including The Institute, plus AASA, the School Superintendents Association, and Day of AI, a non-profit launched out of MIT's RAISE AI Literacy Initiative.

5:01Your charge is pretty simple. This is an opportunity for you to tell us what you think. The students were reminded whatever they come up with wouldn't become real law. After all, they're not real senators. But if they could agree, their bill would be sent to AASA's network of some 10 ,000 school leaders across the country, many of whom are right now crafting their own local AI policies in the absence of national regulation. The adults themselves haven't been able to come to consensus on this. Jeffrey Riley is executive director of Day of AI. Now it's false to the kids. Let them lead the way. I think they recognize that AI is a powerful technology with some really positives, but I think they recognize there's some negatives as well.

5:44After being sworn in, the students were then divided into subcommittees to turn a laundry list of proposed ideas into one coherent bill. The person in writing is states should require schools to provide AI literacy when students first start utilizing technology in the classroom. Arkansas' Danielle Robertson and her team all agreed that schools should be required to teach students about AI and how to use it responsibly, including plagiarism, data privacy, and understanding that chatbots are designed to tell you what you want to hear. But how early should that teaching start? Elijah Martinez, representing New Mexico, argued the earlier the better.

6:25If we don't teach kids what these AI models are using their data for, I think it's dangerous because we're allowing kids to put in their information, and that's private sensitive information. Ultimately, the subcommittee agreed schools should be required to teach kids about AI as soon as they introduce devices into the classroom. Next up, Bina Mejia Romero of Connecticut had to defend this particular amendment in front of the full Senate. We should be introduced to it when devices are introduced to our classrooms. If we're going to be using devices, we should know the dangers and the benefits that they have.

7:03Fantastic. Let's get a round of applause for the senator tonight again. For older students, the senators decided to ban the use of AI on any graded tests and in the act of writing, though editing and studying with AI would be allowed after eighth grade. And for students accused of breaking those rules, Ohio's David McCann-Linares told the full Senate there should be protections. There's been many instances where students have had their work falsely flagged and they haven't been able to appeal it, or they have had to spend multiple weeks appealing it. There must be an appeal process there. Not just an appeal process, but the review should be done by two school adults, not just one.

7:38In another powerful moment on the Senate floor, Dylan Dornack from Iowa told his colleagues that if a student in crisis tries to get emotional support from a school chatbot, the AI should not respond. Instead, it should notify trusted school adults. It'll alert those people, and then those people can go and find that student, have them come and talk to them, provide them with the resources that that student clearly needs in that moment. When asked if alerting adults might betray students' trust, Your friend might be mad at you for telling somebody that they have mentioned self-harm or anything like that, but you might have saved their life.

8:17By the end of the weekend, the final bill had nearly 50 provisions, but still enough inconsistencies and questions that it wasn't clear it would pass. Michael Champany, who works for the Institute, told the students, this is how legislating works. It's messy. And then he reminded them why they were there. There is currently no AI policy across the country. Many schools have nothing in place. If you truly think that nothing is better than what you have put together, you can absolutely vote this bill down. But are you letting perfect be the enemy of good? After a frantic working lunch and some quick fixes, it was time for the final vote.

8:56Dozens of students signed up to make one-minute speeches before the full Senate. This morning, I was one of this bill's staunchest opponents. Ethan Liu traveled to Boston from West Virginia. I had a whole different speech planned out against this bill, actually. It was really good. I was going to quote Charlie XCX, and it was really good. And I'm sad I wasn't able to give it. Instead, Lew encouraged his Senate colleagues to vote yes. After a few more speeches, Time for voting has come, Senators. the student senators cast their votes Senator Caldwell. Senator votes no. Senator Cox. Senator votes aye.

9:31and held their breath. Senators, by a final vote of 83 in favor, 15 opposed. So senators, congratulations. You've passed S-2026, the Students' First Act, here today. Well done. Now, the bill heads to those 10 ,000 school leaders across the country, many of whom are debating these very questions, and who might find the Students' First Act a really good first step. Corey Turner, NPR News, Boston. This episode was produced by Lauren McGocky, Lena Muhammad, and Vincent Acovino, with audio engineering by Ted Meebane. Our director is Jonas Adams. It was edited by Nervie Shaw and Tenbeat Ermias. Our interim executive producer is Courtney Dorning.

10:13It's Consider This from NPR. I'm Juana Summers. This message comes from Workday, the enterprise AI platform for HR, finance, and IT. Since 1949, medical technology company Medtronic has set out to provide life-saving equipment for patients. Now, their workforce has the ultimate tool, more time to focus on those life-saving products. Before, Medtronic recruiters found their days were consumed with managing around 12 ,000 hires a year. With Workday's AI agents for HR, teams can automate interview scheduling and reduce processes from days to just minutes. With 80 % of their HR processes safely automated, recruiters at Medtronic spend less time on routine tasks and more time making strategic decisions, like hiring the next innovators in healthcare.

11:05They do it all through Workday's AI platform, built with a deep understanding of their organization's unique rules and guardrails, ensuring every AI action has the permissions to get work done and get it done right. Learn more at Workday.com.