OpenAI is amending its hastily arranged deal to supply artificial intelligence to the US Department of War (DoW) after the CEO of owner ChatGPT admitted it seemed “opportunistic and careless”.
The contract raised fears that the San Francisco startup’s AI could be used for domestic mass surveillance, but its boss, Sam Altman, said late Monday that the startup would explicitly prohibit its technology from being used for that purpose or used by defense department intelligence agencies like the National Security Agency (NSA).
OpenAI, which has more than 900 million ChatGPT users, closed the deal almost immediately after the Pentagon’s current AI contractor, Anthropic, was dropped.
Anthropic had insisted that “the use of these systems for mass domestic surveillance is incompatible with democratic values,” prompting US President Donald Trump to call Anthropic “left-wing crazy jobs” and order the federal government to stop using its technology.
Despite OpenAI’s denials that the deal allowed the use of surveillance, commentators raised the specter of the Snowden scandal, which broke in 2013, when it emerged that the NSA was involved in the mass collection of telephone and Internet communications.
The deal sparked an online backlash against OpenAI, with X and Reddit users encouraging a “ChatGPT takedown” campaign. One post said: “You are now training a war machine. Let’s see proof of the cancellation.”
Claude, the chatbot created by Anthropic, jumped to the top of Apple’s App Store charts, surpassing ChatGPT, according to an analysis by Sensor Tower.
In a message to employees reposted on X, OpenAI’s CEO said the original deal announced Friday had closed too quickly after Anthropic’s exit.
“We shouldn’t have rushed this out on Friday,” Altman wrote. “The issues are very complex and require clear communication. We were really trying to reduce tensions and avoid a much worse outcome, but I think it seemed opportunistic and careless.”
In announcing the deal, OpenAI had said the contract had “more guardrails than any previous agreement for classified AI deployments, including Anthropic’s.”
However, the US military’s use of AI has alarmed nearly 900 employees at OpenAI and Google, also a leading powerhouse in the technology, who signed an open letter asking their bosses to refuse to allow the War Department to use their products for surveillance and autonomous assassination.
Warning that the US government was trying to “divide each company for fear that the other would give in,” they wrote: “We hope that our leaders will put aside their differences and stand together to continue to reject the War Department’s current demands for permission to use our models for domestic mass surveillance and the autonomous killing of people without human supervision.”
The letter has been signed by 796 Google employees and 98 OpenAI employees. OpenAI said in a blog post announcing the deal with DoW that one of its red lines was “not to use OpenAI technology to direct autonomous weapons systems.”
However, observers, including OpenAI’s former head of policy research Miles Brundage, have questioned how OpenAI has managed to secure a deal that alleviates ethical concerns that Anthropic considered insurmountable. In a post on
Brundage added: “To be clear, the OAI is a complex organization, and I think many people involved in this worked hard to achieve what they consider to be a fair outcome. Some others I don’t trust at all, particularly when it comes to government relations and politics.”
In his X post, he also wrote that he would “rather go to jail” than follow an unconstitutional government order.
“We want to work through democratic processes,” Brundage wrote. “It should be the government that makes key decisions about society. We want to have a voice and a seat at the table where we can share our experience and fight for the principles of freedom.”
Meanwhile, three other U.S. cabinet-level agencies — the departments of State, Treasury, and health and human services — have taken steps to stop using Anthropic’s AI products after the War Department declared the company a supply chain risk. Trump has ordered all US government agencies to phase out use of Anthropic following Defense Secretary Pete Hegseth’s decision.
