Texas Rep. Greg Casar discusses bill to temporarily pause advanced AI development

NPR's Leila Fadel speaks with U.S. Rep. Greg Casar, D-Texas, about his efforts to regulate artificial intelligence.

LEILA FADEL, HOST: Let's turn now to efforts in Congress to rein in artificial intelligence. One of the leading voices in that effort is Congressman Greg Casar. He's a progressive Democrat from Texas. Good morning, Congressman, and welcome back to the program. GREG CASAR: Good morning. FADEL: So you're introducing legislation with your Senate colleague Bernie Sanders that would ban, quote, "superintelligent AI." What is superintelligent AI, and why do you feel that it's such a danger? CASAR: You can think of superintelligent AI as like an AI superweapon - artificial intelligence that becomes smarter than humans, impossible for us to control, able to defy commands to be turned off and that could have the capability of developing nuclear weapons or other weapons of mass destruction. And so Senator Sanders' and my bill isn't bold. It's just actually common sense. No company should be able to build this kind of artificial intelligence that humans could not control or contain. FADEL: Is there support for your bill? CASAR: Data for Progress, a progressive think tank, recently pulled the country and actually found that about 70% of Americans support an immediate ban on artificial superintelligence, with only about 25% opposed. Americans every single day are hearing more and more about the dangers of completely unregulated AI. FADEL: But is there support for your bill among your colleagues, I mean, to get this passed? CASAR: I have found incredible support just talking about it. The bill text will actually be released in the coming days. And so this is one of the few bills I've experienced in Congress where we haven't even released the bill yet and are already getting significant amounts of support because this is just a baseline national security issue. FADEL: Now, President Trump is dismissing the dangers of AI and warning that if there is a slowdown, then the U.S. loses the AI race to China. Is he wrong? Will being cautious lose the U.S. their competitive edge? CASAR: Donald Trump is wrong here. We must pursue international agreements, including with China, to prevent extremely dangerous AI from being developed anywhere in the world. We can do that. We've done that during the Cold War when we signed nuclear nonproliferation agreements with the Soviet Union. The U.S. and China are in this race. But it's the kind of race where the second you step over the finish line, you fall down a steep, deadly cliff. No one can win the AI arms race. Our goal has to be to end the AI arms race to protect people here and around the world. FADEL: Now, some of the scenarios experts are warning about is that humans will lose control of the technology, and that could hypothetically lead to mass extinction. So this is a global threat. Now, let's say your bill passes. That doesn't regulate anything globally. It just says, pursue agreements. I mean, how do you work in conjunction with other world powers, namely China, to make sure this is a global effort? CASAR: We need to put up AI safety guardrails now in the United States to protect Americans while we pursue international agreements. By putting in protections here in the United States, we can show we're serious about getting a deal. And because Chinese models currently train on American AI models, passing strong U.S. rules now along with export controls slow down dangerous AI development around the world. And so we should make our country lead, of course, in technology but also be the leader in the world on AI diplomacy. This should be a top priority for the United Nations meeting next week. America should be laying the groundwork for international agreements to block superdangerous AI. FADEL: Now, if the president is not in agreement with you, do you want Congress to intervene on its own? CASAR: Congress absolutely should intervene, and Mike Johnson should be making this a top priority. But unfortunately, our speaker - Republican Speaker Mike Johnson - said that it's big tech companies that should be in charge of AI safety. That's like saying the bank robbers should be in charge of law enforcement. Elected officials in Congress should be cracking down on AI companies that are breaking laws and admitting their huge safety threats. But instead, we see our Republican leadership, from Mike Johnson to Donald Trump, courting these AI billionaires for gigantic campaign contributions from their super PACs. FADEL: I mean, were you surprised, though, that over the weekend, a number of leaders in the industry - Anthropic, OpenAI, Google - said they supported a coordinated industry slowdown in development? CASAR: Their CEOs are talking about a pause. They're talking about something voluntary, but voluntary is the whole problem. Every one of these companies just making a pinky promise they can revoke is not a safety system. That's why Congress needs to act, and one important step is to pass Senator Sanders' and my bill to have a federal agency that stops the madness and keeps Americans safe first. FADEL: Greg Casar is a Democrat representing Texas' 35th Congressional District. Congressman, thank you. CASAR: Thank you. FADEL: And we will note - Anthropic and Google are financial supporters of NPR. Copyright © 2026 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information. Accuracy and availability of NPR transcripts may vary. Transcript text may be revised to correct errors or match updates to audio. Audio on npr.org may be edited after its original broadcast or publication. The authoritative record of NPR’s programming is the audio record.

LEILA FADEL, HOST: Let's turn now to efforts in Congress to rein in artificial intelligence. One of the leading voices in that effort is Congressman Greg Casar. He's a progressive Democrat from Texas. Good morning, Congressman, and welcome back to the program. GREG CASAR: Good morning. FADEL: So you're introducing legislation with your Senate colleague Bernie Sanders that would ban, quote, "superintelligent AI." What is superintelligent AI, and why do you feel that it's such a danger? CASAR: You can think of superintelligent AI as like an AI superweapon - artificial intelligence that becomes smarter than humans, impossible for us to control, able to defy commands to be turned off and that could have the capability of developing nuclear weapons or other weapons of mass destruction. And so Senator Sanders' and my bill isn't bold. It's just actually common sense. No company should be able to build this kind of artificial intelligence that humans could not control or contain. FADEL: Is there support for your bill? CASAR: Data for Progress, a progressive think tank, recently pulled the country and actually found that about 70% of Americans support an immediate ban on artificial superintelligence, with only about 25% opposed. Americans every single day are hearing more and more about the dangers of completely unregulated AI. FADEL: But is there support for your bill among your colleagues, I mean, to get this passed? CASAR: I have found incredible support just talking about it. The bill text will actually be released in the coming days. And so this is one of the few bills I've experienced in Congress where we haven't even released the bill yet and are already getting significant amounts of support because this is just a baseline national security issue. FADEL: Now, President Trump is dismissing the dangers of AI and warning that if there is a slowdown, then the U.S. loses the AI race to China. Is he wrong? Will being cautious lose the U.S. their competitive edge? CASAR: Donald Trump is wrong here. We must pursue international agreements, including with China, to prevent extremely dangerous AI from being developed anywhere in the world. We can do that. We've done that during the Cold War when we signed nuclear nonproliferation agreements with the Soviet Union. The U.S. and China are in this race. But it's the kind of race where the second you step over the finish line, you fall down a steep, deadly cliff. No one can win the AI arms race. Our goal has to be to end the AI arms race to protect people here and around the world. FADEL: Now, some of the scenarios experts are warning about is that humans will lose control of the technology, and that could hypothetically lead to mass extinction. So this is a global threat. Now, let's say your bill passes. That doesn't regulate anything globally. It just says, pursue agreements. I mean, how do you work in conjunction with other world powers, namely China, to make sure this is a global effort? CASAR: We need to put up AI safety guardrails now in the United States to protect Americans while we pursue international agreements. By putting in protections here in the United States, we can show we're serious about getting a deal. And because Chinese models currently train on American AI models, passing strong U.S. rules now along with export controls slow down dangerous AI development around the world. And so we should make our country lead, of course, in technology but also be the leader in the world on AI diplomacy. This should be a top priority for the United Nations meeting next week. America should be laying the groundwork for international agreements to block superdangerous AI. FADEL: Now, if the president is not in agreement with you, do you want Congress to intervene on its own? CASAR: Congress absolutely should intervene, and Mike Johnson should be making this a top priority. But unfortunately, our speaker - Republican Speaker Mike Johnson - said that it's big tech companies that should be in charge of AI safety. That's like saying the bank robbers should be in charge of law enforcement. Elected officials in Congress should be cracking down on AI companies that are breaking laws and admitting their huge safety threats. But instead, we see our Republican leadership, from Mike Johnson to Donald Trump, courting these AI billionaires for gigantic campaign contributions from their super PACs. FADEL: I mean, were you surprised, though, that over the weekend, a number of leaders in the industry - Anthropic, OpenAI, Google - said they supported a coordinated industry slowdown in development? CASAR: Their CEOs are talking about a pause. They're talking about something voluntary, but voluntary is the whole problem. Every one of these companies just making a pinky promise they can revoke is not a safety system. That's why Congress needs to act, and one important step is to pass Senator Sanders' and my bill to have a federal agency that stops the madness and keeps Americans safe first. FADEL: Greg Casar is a Democrat representing Texas' 35th Congressional District. Congressman, thank you. CASAR: Thank you. FADEL: And we will note - Anthropic and Google are financial supporters of NPR. Copyright © 2026 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information. Accuracy and availability of NPR transcripts may vary. Transcript text may be revised to correct errors or match updates to audio. Audio on npr.org may be edited after its original broadcast or publication. The authoritative record of NPR’s programming is the audio record.

Let's turn now to efforts in Congress to rein in artificial intelligence. One of the leading voices in that effort is Congressman Greg Casar. He's a progressive Democrat from Texas. Good morning, Congressman, and welcome back to the program. GREG CASAR: Good morning. FADEL: So you're introducing legislation with your Senate colleague Bernie Sanders that would ban, quote, "superintelligent AI." What is superintelligent AI, and why do you feel that it's such a danger? CASAR: You can think of superintelligent AI as like an AI superweapon - artificial intelligence that becomes smarter than humans, impossible for us to control, able to defy commands to be turned off and that could have the capability of developing nuclear weapons or other weapons of mass destruction. And so Senator Sanders' and my bill isn't bold. It's just actually common sense. No company should be able to build this kind of artificial intelligence that humans could not control or contain. FADEL: Is there support for your bill? CASAR: Data for Progress, a progressive think tank, recently pulled the country and actually found that about 70% of Americans support an immediate ban on artificial superintelligence, with only about 25% opposed. Americans every single day are hearing more and more about the dangers of completely unregulated AI. FADEL: But is there support for your bill among your colleagues, I mean, to get this passed? CASAR: I have found incredible support just talking about it. The bill text will actually be released in the coming days. And so this is one of the few bills I've experienced in Congress where we haven't even released the bill yet and are already getting significant amounts of support because this is just a baseline national security issue. FADEL: Now, President Trump is dismissing the dangers of AI and warning that if there is a slowdown, then the U.S. loses the AI race to China. Is he wrong? Will being cautious lose the U.S. their competitive edge? CASAR: Donald Trump is wrong here. We must pursue international agreements, including with China, to prevent extremely dangerous AI from being developed anywhere in the world. We can do that. We've done that during the Cold War when we signed nuclear nonproliferation agreements with the Soviet Union. The U.S. and China are in this race. But it's the kind of race where the second you step over the finish line, you fall down a steep, deadly cliff. No one can win the AI arms race. Our goal has to be to end the AI arms race to protect people here and around the world. FADEL: Now, some of the scenarios experts are warning about is that humans will lose control of the technology, and that could hypothetically lead to mass extinction. So this is a global threat. Now, let's say your bill passes. That doesn't regulate anything globally. It just says, pursue agreements. I mean, how do you work in conjunction with other world powers, namely China, to make sure this is a global effort? CASAR: We need to put up AI safety guardrails now in the United States to protect Americans while we pursue international agreements. By putting in protections here in the United States, we can show we're serious about getting a deal. And because Chinese models currently train on American AI models, passing strong U.S. rules now along with export controls slow down dangerous AI development around the world. And so we should make our country lead, of course, in technology but also be the leader in the world on AI diplomacy. This should be a top priority for the United Nations meeting next week. America should be laying the groundwork for international agreements to block super