The Pentagon keeps promising to follow the law when using AI, but what are the limits?

The Pentagon has been using AI tools such as Anthropic's Claude to assist in military targeting by sifting through data to flag potential targets for commanders. However, current law does not place explicit limits on where AI can be used in the kill chain, raising questions about accountability and oversight.

Despite these promises, the absence of specific legal boundaries means that AI could theoretically be involved in more steps of the targeting decision process without clear regulation. The use of Claude for flagging targets is already underway.

The Pentagon's stated approach is to maintain 'human oversight' over AI decisions, but critics argue that without legal guardrails, the definition of oversight could erode over time. The U.S. The article does not specify any particular incident of misuse but highlights the gap between existing law and technological capability.