Anthropic CEO Dario Amodei is sounding the alarm on artificial intelligence. The AI leader recently released a 38-page essay outlining escalating risks tied to the technology his company is pioneering, and he shared his urgent recommendations with lawmakers during a conversation in San Francisco.
Amodei laid out a three-part prescription for Congress:
-
Transparency Laws for AI – AI companies must disclose the risks of their models, any harmful behavior, and the safeguards in place to prevent misuse. Amodei emphasized that understanding the limits and vulnerabilities of AI systems is critical for public safety.
-
Restrict AI-Related Technology Exports to China – The CEO urged lawmakers to halt sales of Nvidia chips and other U.S. products that enable AI development abroad, particularly in countries where oversight and ethical controls may be weaker.
-
Prepare to Tax AI Trillionaires – Amodei called for policies to ensure the immense wealth generated by AI is shared responsibly. “You’re going to get a mob coming for you if you don’t do this the right way,” he said, warning potential AI billionaires about the societal consequences of hoarding unprecedented fortunes.
Underlying all three points is a simple principle: anticipate the worst-case scenarios. “We always assume that everything that can go wrong does go wrong,” Amodei said. “That’s how you build things that are reliable.” His message stresses not only caution but proactive governance as AI systems become more capable and widespread.
Amodei’s essay and public statements are part of a growing chorus of industry leaders calling for regulation, transparency, and global cooperation on AI safety. As AI technologies continue to advance at a rapid pace, his recommendations provide lawmakers with concrete steps to reduce risk while harnessing the potential of this transformative technology.
