Dario Amodei needs to face his own duty to prevent AI-mageddon

Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman are calling for federal regulation of frontier AI development, a plan White House tech adviser David Sacks says the companies do not need permission to act on themselves. President Donald Trump and House Speaker Mike Johnson have questioned Amodei's push for government intervention, while Meta CEO Mark Zuckerberg argues against "AI centralization" and favors competition over a global accord that would require cooperation from Beijing. The op-ed contends Amodei and Altman can slow their own frontier work without federal action and have a duty not to cause harm.

“Stop us before we doom humanity” is how tech kingpins Dario Amodei and Sam Altman present what amounts to a call for Uncle Sam to permanently lock in their companies’ leadership in AI. No one much doubts that Anthropic and OpenAI are the cutting-edge artificial-intelligence firms, so if Amodei and Altman https://nypost.com/2026/09/12/us-news/ai-ceos-demand-the-industry-slow-the-pace-of-advancements/ think that their “frontier” work is running into risky territory, they can slow down without need for federal action — and should. President Donald Trump https://nypost.com/2026/09/13/tech/trump-brushes-off-ai-doom-predictions-says-us-must-push-ahead/ and House Speaker Mike Johnson are entirely correct to question https://nypost.com/2026/09/13/us-news/speaker-johnson-says-ai-leaders-must-meet-with-congress-on-safety-laws-after-end-of-humanity-warnings/ Amodei’s call for the feds to “help” them do it. They don’t “need anyone else’s permission,” notes David Sacks, the top White House tech adviser — nor any suspension of antitrust law or other protections against liability law; they certainly don’t need their pet evaluators empowered “to police competitors who aren’t even at the frontier.” An added red herring in Amodei’s Altman-endorsed plan to supposedly avert AI-mageddon is the pretense that some global deal limiting AI progress is possible. Such an accord would require Beijing to play nice — when both companies and every other major firm regularly say China is stealing https://nypost.com/2026/07/13/business/how-china-is-ripping-off-cutting-edge-ai-from-anthropic-openai-and-threatening-us-national-security/ their stuff, e.g. by using Claude to train up its own AIs. This is the country that gave the world COVID via its careless biowarfare research: It’s impossible to trust the Chinese Communist Party to be any more careful when it comes to the tech likely to dominate the future world economy. On this count, Meta’s Mark Zuckerberg has the best approach https://nypost.com/2026/07/29/business/metas-mark-zuckerberg-blasts-ai-centralization-opposes-us-ban-on-chinese-models/ : The world is most secure without “AI centralization.” Competition is the best bet for ensuring no one company’s or country’s work can take over. We don’t mean to accuse Amodei of simply trying to pull a fast one with his plan: Like many other high-IQ types, these tech titans tend to obsess over their own insights, letting a partial truth obscure the bigger picture. Thus Amodei and Altman overlook the fact that they have the power to slow down — and, like any company, a duty to not cause harm. Take responsibility for your own work, gents, and stop pretending Uncle Sam or Big Brother is there to do it for you.