AI for Military Support
Research suggests that despite concerns about AI automating military decisions, human operators exhibit ‘algorithmic aversion’ when presented with AI recommendations, particularly in scenarios with potential for significant collateral damage. However, incorporating ‘explainable AI’ features significantly reduces this aversion and encourages more careful evaluation of AI suggestions, highlighting the continued need for human oversight.
A recent study investigated how artificial intelligence is impacting military decision-making processes. Researchers reconstructed a military targeting AI decision-support system (DSS) and conducted two experiments with 2,015 Israeli military personnel to assess its influence on combat decisions. The study’s findings contradict the common belief that AI would lead to unquestioning acceptance of its recommendations. Instead, the research revealed a phenomenon known as ‘algorithmic aversion,’ where operators actively resisted AI suggestions, especially when the potential for civilian casualties was high. This resistance was not simply random; it appeared to be influenced by individual factors, the perceived importance of the operation, and the design of the AI interface itself. The researchers found that incorporating ‘explainable AI’ – features that provide insights into how the AI reached its conclusions – significantly mitigated algorithmic aversion and encouraged more deliberate consideration of the AI’s recommendations. This research emphasizes the ongoing role of human judgment and the need to balance AI assistance with careful oversight in military contexts.