The US military is increasingly using AI and machine learning technologies to aid in precision targeting of enemy groups in the Middle East.
Computer vision algorithms are being used to identify targets in regions like Yemen and Iraq from a variety of data sources.
AI systems can efficiently analyze large amounts of signals intelligence, social media, and other online data to discern patterns and locate individuals.
This enhances the targeting process by estimating locations of targets and likelihood of their presence, as well as predicting strike success and collateral damage.
While these systems may help refine targeting, there is little transparency into their exact operations and ability to avoid civilian casualties.
Steven Feldstein said, “AI systems can aggregate bulk information and identify targets and locations in ways that can be more efficient than relying on traditional techniques.”
“Given how much digital communication is permeating our ecosystem, the ability of AI systems to sift through the digital noise and discern patterns linked to specific targets can be immensely valuable for these operations,” he continued.
Feldstein explained, “If given access to the right data, systems can comb through signals intelligence, social media posts and other factors that would provide information about a target’s whereabouts, make an estimate as to the likelihood that the target currently resides in a specific location and possibly give an estimate about the potential success of a strike, what kind of strike would best accomplish the objective and potential collateral damage from such a strike.”
Retired Rear Adm. Mark Montgomery, senior director of the Center on Cyber and Technology Innovation at the Foundation for the Defense of Democracies, said, “As AI tools come online that enhance this machine learning process, AI has become an element of the U.S. targeting process as well. This is a natural extension of the high-quality process U.S. intelligence and targeting officials utilize.”
Feldstein concluded, “There isn’t a lot of transparency about how AI targeting systems work and whether they are able to determine a risk of civilian casualties from a particular strike. Until we have more information about parameters, accountability of use and other related factors, we simply have no reason to know (nor can we assume) that these systems help avoid civilian casualties.”
Some experts argue AI has become an integral part of US targeting but more information is needed to understand accountability and safety parameters.
The use of these advanced technologies raises questions about transparency as the military ramps up airstrikes against threats like Houthi and Iraqi militant groups.
