Anthropic CEO says proposed 10-year ban on state AI regulation ‘too blunt’ in NYT op-ed

Audio By Carbonatix

Enlarge

(Reuters) -A Republican proposal to block states from regulating artificial intelligence for 10 years is “too blunt,” Anthropic Chief Executive Officer Dario Amodei wrote in a New York Times’ opinion piece.

Amodei instead called for the White House and Congress to work together on a transparency standard for AI companies at a federal level, so that emerging risks are made clear to the people.

“A 10-year moratorium is far too blunt an instrument. AI is advancing too head-spinningly fast,” Amodei said.

“Without a clear plan for a federal response, a moratorium would give us the worst of both worlds – no ability for states to act, and no national policy as a backstop.”

The proposal, included in President Donald Trump’s tax cut bill, aims to preempt AI laws and regulations passed recently in dozens of states, but has drawn opposition from a bipartisan group of attorneys general that have regulated high-risk uses of the technology.

Instead, a national standard would require developers working on powerful models to adopt policies for testing and evaluating their models and to publicly disclose how they plan to test for and mitigate national security and other risks, according to Amodei’s opinion piece.

Such a policy, if adopted, would also mean developers would have to be upfront about the steps they took to make sure their models were safe before releasing them to the public, he said.

Amodei said Amazon.com-backed Anthropic already releases such information and competitors OpenAI and Google DeepMind have adopted similar policies.

Legislative incentives to ensure that these companies keep disclosing such details could become necessary as corporate incentive to provide this level of transparency might change in light of models becoming more powerful, he argued.

(Reporting by Arsheeya Bajwa in Bengaluru; Editing by Anil D’Silva)