Signed a classified contract on the same day, Google quietly withdrew from the Pentagon's drone swarm competition

According to Singularity Beat monitoring, Google participated in a $100 million Pentagon autonomous drone swarm technology competition at the end of January and successfully qualified, but notified the government of its withdrawal a few weeks later. The decision to withdraw was due to internal ethical reviews, with the public explanation citing "resource constraints." Several Google employees involved in the project expressed disappointment at the withdrawal.

The competition was jointly led by the Special Operations Command's Defense Autonomous Operations Group and the Defense Innovation Unit, with the goal of allowing commanders to control drone swarms through voice commands. The competition lasted for six months, with subsequent stages involving "target-centric perception and sharing" and "from launch to effects." OpenAI, Palantir, and xAI have been selected to continue participating. Anthropic also submitted a proposal but was not selected, as the company's assessment determined that the proposal did not align with CEO Dario Amodei's set boundary of "not engaging in fully autonomous weapon" projects.

Google's conflicting stance was starkly exposed on the same day: while withdrawing from the drone swarm competition, the company revised its existing contracts with the Pentagon to allow the military to use its AI for "any and all legal government purposes," including classified work. A Google spokesperson stated that this was simply a routine corporate client adjustment for consumer-grade security filters, "not involving the development of custom models for the Pentagon," and emphasized that "AI should not be used for autonomous weapons without proper human oversight."