In today’s AI news, Anthropic CEO Dario Amodei is sounding the alarm over a proposal in President Trump’s latest spending bill that would bar states from regulating artificial intelligence for ten years. Writing in the New York Times, Amodei warns, “A 10-year moratorium is far too blunt an instrument. A.I. is advancing too head-spinningly fast.” He argues that we could see fundamental changes to our world in as little as two years, and by a decade from now, “all bets are off.”

Amodei cautions that without both state action and a clear federal plan, such a ban would leave us with “no ability for states to act, and no national policy as a backstop.” Instead, he urges Congress and the White House to collaborate on new AI transparency standards so emerging risks are laid bare for the American public.

Meanwhile, Anthropic is facing its first major legal challenge: Reddit has sued the company for using user content to train its models without permission. Reddit’s chief legal officer, Ben Lee, says, “We will not tolerate profit-seeking entities like Anthropic commercially exploiting Reddit content for billions of dollars without any return for redditors or respect for their privacy.”
Link to Article