The U.S. military is increasingly using artificial intelligence to help identify potential targets in ongoing air operations involving Iran, according to two people familiar with the matter who spoke on condition of anonymity. Reports Nbcnews.com.
The sources said the military is using AI-enabled software from Palantir, including its Maven intelligence analysis program. Maven relies in part on Anthropic’s Claude systems to help analysts sift through large volumes of intelligence, though a person familiar with the work said Claude does not directly generate targeting advice.
The expanded role of AI is drawing scrutiny on Capitol Hill. Rep. Jill Tokuda of Hawaii, a member of the House Armed Services Committee, called for an “impartial review” to assess whether AI use has already put lives at risk, stressing that human judgment must remain central in life-or-death decisions.
Rep. Sara Jacobs of California warned that AI tools can fail in subtle ways and may be over-trusted by operators, urging strict guardrails and a guaranteed “human in the loop” for any lethal action.
The Pentagon has publicly said it does not want autonomous weapons operating without human involvement, but it has not detailed how it verifies AI-driven analysis. The issue is further complicated by tensions between Defense Secretary Pete Hegseth and Anthropic over limits on military uses, including concerns about domestic surveillance and autonomous lethal systems.
Read “Zamin” on Telegram!
