'Needless display of brute punishment': WSJ editorial rips into Trump's war with AI firm

The Wall Street Journal editorial board sharply condemned President Donald Trump on Friday for ordering the federal government to blackball Anthropic, an artificial intelligence firm that had refused to relax its safety rules to let the Defense Department deploy automated surveillance and military strikes.

Trump proclaimed in a Truth Social rant earlier in the day that the "Leftwing nut job" firm, which operates the Claude AI model, was forcing the Pentagon to "obey their Terms of Service instead of our Constitution" and vowed that all federal contracts with the company would be phased out.

"The Pentagon is within its rights to stop working with the company," wrote the board. "The missions of the U.S. military are the responsibility of elected and politically accountable officials. It’s an imperfect analogy, but a company can’t sell the U.S. military a missile and then haggle about acceptable targets."

Nonetheless, the board continued, Trump's flat-out attack on the company is a "needless display of brute government punishment" that goes way too far — and could chill the entire American AI industry.

"Anthropic doesn’t lack for patriotism. The company says it has left revenue on the table by cutting off firms linked to the Chinese Communist Party. It’s no small matter that a technology company has been willing to help the U.S. military in combat, a change from a decade ago when most of Silicon Valley viewed Pentagon contracts as complicity in imperialism," wrote the board. Indeed, "the bigger picture before the meltdown was that an AI company with a progressive reputation and the Trump Pentagon largely agreed that America has to be defended with premiere technology."

The bottom line, wrote the board, is that "The Pentagon needs all the AI help it can get as the technology races ahead and China isn’t far behind. The People’s Liberation Army is the winner of the Anthropic ban."