OpenAI CEO Sam Altman told employees on Tuesday that the company does not control how the Department of Defense uses its artificial intelligence products in military operations. Altman’s assertion that OpenAI is lacking input comes amid increased scrutiny of how the military uses AI in warfare and growing ethical concerns from AI workers about how the technology is deployed.
“I can’t make business decisions,” Altman told employees, Bloomberg and CNBC reported.
“So you might think attacking Iran was good and invading Venezuela was bad. You don’t have to consider that,” Altman was quoted as saying.
The AI industry has been mired in heated debate and tough negotiations in recent weeks, as the Department of Defense requires AI companies to remove safety guardrails from their models to allow broader military applications. AI-enabled systems are reportedly already being used in the U.S. military’s operation to capture Venezuelan leader Nicolas Maduro and in determining targets in the war against Iran.
OpenAI rival Anthropic, maker of the Claude chatbot, last week rejected a contract with the Department of Defense over concerns that its model could be used for domestic mass surveillance or fully autonomous weapons. As a result, U.S. Secretary of Defense Pete Hegseth declared the company at “supply chain risk,” a designation that has never been used against U.S. companies and could cause significant economic harm if formalized.
On the same day that Hegseth vowed punitive action against Anthropic, the Pentagon also announced a contract with OpenAI, apparently intended to replace the use of Claude in military applications. The timing of the deal and concerns that OpenAI had agreed to cross an ethical line that Anthropic refused to do led to both public and internal employee backlash against OpenAI.
Since then, Altman and OpenAI have tried to argue that their technology was used legally and to do damage control, and Altman acknowledged that the deal was rushed and made the company look “opportunistic and sloppy.”
In a memo to employees on Wednesday, Anthropic CEO Dario Amodei called Mr. Altman “shady” and accused him of “admiring Mr. Trump like a dictator,” the Information reported. “In fact, rather than colluding with our employees to create a ‘safety theater’ for the benefit of our employees, we have sincerely followed the lines that should not be crossed (which, I swear, is something literally everyone is working on) [the Pentagon]Palantir, our political consultants, and others assumed that was the problem we were trying to solve),” Amodei reportedly wrote.
He also took aim at the Pentagon and Donald Trump.
“The real reason is [the Pentagon] And what the Trump administrators don’t like about us is that we’re not donating to Trump (while OpenAI/Greg is giving a lot of money),” he said, referring to OpenAI president Greg Brockman, who along with his wife donated $25 million to the pro-Trump pack.
