Google has signed a new contract expanding the US Department of Defense's access to its AI systems, stepping in after Anthropic declined to permit the DoD to use its models for domestic mass surveillance and autonomous weapons systems.
Catch: The specific terms, scope, and safeguards of Google's DoD contract have not been publicly disclosed, making independent verification of limits difficult.
Flip side: Anthropic's refusal demonstrates that AI labs can and do exercise ethical restraint, setting a precedent that other companies could follow.