Key facts
- Rep. Greg Casar believes the AI race is not one to win and advocates for international agreements to prevent dangerous AI development.
- Casar is co-sponsoring a bill with Sen. Bernie Sanders to ban "superintelligent AI," defined as AI smarter than humans and uncontrollable.
- A recent poll indicates about 70% of Americans support an immediate ban on artificial superintelligence.
- President Trump has dismissed AI dangers, warning that a slowdown would cede technological supremacy to China.
- Casar argues that strong U.S. regulations and export controls can slow dangerous AI development globally.
- Casar wants the U.S. to lead AI diplomacy and establish international agreements on AI safety.
Representative Greg Casar is advocating for a global approach to AI safety, arguing that the pursuit of an "AI arms race" is dangerous and ultimately unwinnable. Speaking at a conference organized by the Future of Life Institute, Casar urged activists and whistleblowers to speak out more forcefully about the risks associated with advanced AI, stating that "tempered rhetoric has passed."
Casar, a progressive Democrat from Texas, is co-sponsoring legislation with Senator Bernie Sanders that aims to ban the development of "superintelligent AI" – artificial intelligence that surpasses human control and could potentially develop weapons of mass destruction. He believes that such AI is an "AI superweapon" and that no company should be allowed to build it.
Recent polling by Data for Progress indicates significant public support for such measures, with about 70% of Americans favoring an immediate ban on artificial superintelligence. This sentiment reflects growing anxieties about the technology's capabilities, with a POLITICO Poll showing two out of three Americans perceive at least a "moderate" risk of advanced AI destroying humanity.
Casar's stance contrasts with that of President Donald Trump, who has dismissed concerns about AI dangers and warned that slowing down development would allow China to gain a competitive edge. Casar argues that Trump is wrong and that international agreements, including with China, are necessary to prevent the development of extremely dangerous AI globally. He draws a parallel to nuclear nonproliferation agreements signed during the Cold War.
To address these concerns, Casar proposes that the U.S. should implement AI safety guardrails domestically while simultaneously pursuing international agreements. He believes that by establishing strong U.S. rules and export controls, the country can influence global AI development, particularly as Chinese models train on American AI. Casar advocates for the U.S. to lead in AI diplomacy, making it a top priority for international forums like the United Nations General Assembly.