OpenAI CEO Sam Altman told employees Tuesday that his company does not control how the Pentagon uses its artificial intelligence products in military operations. Altman’s claims about OpenAI’s lack of input come amid increased scrutiny over how the military uses AI in warfare and ethical concerns from AI workers about how their technology will be deployed.
“You can’t make operational decisions,” Altman told employees, according to reports from Bloomberg and CNBC.
“So maybe you think the attack on Iran was good and the invasion of Venezuela was bad. You can’t have an opinion on that,” Altman allegedly said.
The AI industry has been embroiled in heated discussions and acrimonious negotiations in recent weeks, as the Pentagon has required AI companies to remove guardrails from their models to enable a broader range of military applications. AI-enabled systems have reportedly already been used in the US military operation to capture Venezuelan leader Nicolás Maduro and in targeting decisions in its war against Iran.
Anthropic, OpenAI’s rival and creator of the Claude chatbot, last week rejected a deal with the Pentagon over fears its model could be used for domestic mass surveillance or fully autonomous weapons. As a result, US Defense Secretary Pete Hegseth declared the company a “supply chain risk,” a designation never before used against a US company and one that could cause significant financial damage if formally enacted.
On the same day that Hegseth promised punitive measures against Anthropic, the Pentagon also announced a deal with OpenAI that was apparently aimed at replacing the use of Claude in military applications. The timing of the deal and concerns that OpenAI had agreed to cross ethical lines that Anthropic rejected sparked a backlash from both the public and internal employees against OpenAI.
Since then, Altman and OpenAI have tried to insist that their technology will be used legally and damage controlled, with Altman admitting that the deal was hastily made and made the company look “opportunistic and careless.”
Anthropic CEO Dario Amodei blasted Altman as a “liar” in a memo to employees Wednesday and accused Altman of giving “dictatorial-style praise to Trump,” Information reported. “In fact, we have maintained our red lines with integrity instead of colluding with them to produce ‘security theater’ for the benefit of employees (which, I absolutely swear, is what literally everyone at (the Pentagon), Palantir, our political consultants, etc., assumed was the problem we were trying to solve),” Amodei allegedly wrote.
He also took aim at the Pentagon and Donald Trump.
“The real reasons they (the Pentagon) and the Trump admin don’t like us is that we haven’t donated to Trump (while OpenAI/Greg have donated a lot),” he wrote, referring to Greg Brockman, president of OpenAI, who gave a Trump-supporting Pac $25 million along with his wife.
