OpenAI CEO Sam Altman revealed today that the company has reached an agreement with the US Department of War (DoW) to integrate its advanced AI models into the department’s classified networks. The announcement, made via a post on X (formerly Twitter), highlights a shared commitment to AI safety and ethical use, while addressing concerns over potential misuse in military applications – a point of friction previously between the US government and Anthropic.
In his statement, Altman highlighted the DoW’s “deep respect for safety and a desire to partner to achieve the best possible outcome.” He reiterated OpenAI’s core mission of promoting AI safety and ensuring the broad distribution of its benefits. Key principles outlined in the agreement include prohibitions on domestic mass surveillance and the requirement for human oversight in decisions involving the use of force, particularly with autonomous weapon systems.
“The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement,” Altman wrote. The deal also incorporates technical safeguards to ensure model behaviour aligns with ethical standards, with deployments limited to cloud networks. Additionally, OpenAI plans to station Frontier Defense Experts (FDEs) to oversee implementation and maintain safety protocols.
Altman’s OpenAI joins hand with US DoW
Altman called on the DoW to extend identical terms to other AI firms, expressing optimism that such arrangements could de-escalate tensions from “legal and governmental actions” toward collaborative solutions. “We remain committed to serve all of humanity as best we can,” he concluded, acknowledging the complexities of a “complicated, messy, and sometimes dangerous” world.
However, the announcement sparked a wave of backlash, with critics accusing OpenAI of compromising on ethics to secure a lucrative government contract, especially after the episode involving Anthropic and the US DoW over the same concern.
Internet backlash over deal with US Govt
The announcement comes amid apparent friction in the AI industry, with several responses referencing competitor Anthropic. One user, @DrogoEth, urged Altman to “learn something from Anthropic,” accompanied by an image implying ethical superiority. Another, @ChristopherHale, declared he had canceled his ChatGPT subscription in favour of Anthropic’s Claude Pro Max, stating, “One stands up for the God-given rights of the American people. The other folds to tyrants.”
AI safety researcher @StephenLCasper questioned the consistency of the DoW’s standards, suggesting Altman might be “misleading” in his portrayal. “Either the DoW has different standards for OAI vs Anthropic, or Altman is being misleading in this tweet. Given his history, I’d expect the latter,” Casper posted.
Other reactions ranged from outright condemnation, like calling Altman “a stain to humanity” to satirical memes, such as one from @shitpost9000 labelling him “the man who sold out humanity.”
The deal comes on the heals of Anthropic turning away from the US DoW’s offer, wherein CEO Dario Amodei was opposed to the provisions on domestic mass surveillance and use in autonomous weapons.