Artificial Intelligence and Machine Learning (AI/ML) have the potential to significantly increase the efficiency and effectiveness of military actions, yet there appears to be a gap in knowledge and action when it comes to acquiring and implementing these tools on a large scale for mission sets like information operations (IO). What AI-enabled suite of tools could enable the Information Warfare Numbered Air Force (NAF) to increase the pace and quality of Information Operations?
To maximize the contributions of both human and AI teammates, the military must develop a specific vision for AI-enabled tool suites that integrate into existing planner and operator processes. Because AI/ML is already used worldwide in programming and malware development, is it possible to harness this capability to rapidly develop applications and scripts that aid cyber operators in performing complicated offensive and defensive operations on demand and at scale in rapidly changing environments? Such an AI/ML solution—equipped with access to a wide variety of information—could provide a more comprehensive assessment of a problem than a human, rendering a list of possible solutions and rationales that cyber operators could consider for action. For example, how can AI/ML-assisted weaponeering and apportionment provide operators with probabilities of success, alternate or supplemental actions, and recommended actions? Furthermore, how could it simplify or automate capability pairing and battle damage assessment (BDA) actions through target identification, critical node assessments, evaluating weapon effectiveness or critical infrastructure vulnerability, and establishing indicators of non-kinetic success or inflection points for defensive action?
As the military seeks to modernize these U.S. cyber-enabled information operations, what are the critical policy and technical limitations to harnessing AI and ML tools, and what are the key requirements for solutions to overcome them? Specifically, what legal and ethical considerations (e.g., human-in-the-loop, bias) exist when integrating AI/ML into the development process? Ultimately, what process and technical controls should be implemented to maintain positive control over the application and the effects generated by its use?
- Banner, Maj. Jeffery A., "Hunt Forward Operations as a Diplomatic Arm of the United States," AF Fellow Op-Ed (University of Texas, San Antonio, Cyber Warfare), 2025, 3 pgs.
- Banner's 2025 Defense Fellow research directly tackles this possibility. Working alongside DoD and academic institutions, his research focuses on automating cyber vulnerability analysis by incorporating artificial intelligence to create "proof of concept exploits and defensive alerts." He notes that this AI integration is designed specifically to better support the capabilities of both offensive and defensive U.S. cyber operators.
- Pantaleon, Maj. Bridget, "Are We Prepared for When Artificial Intelligence Costs Innocent Lives?" AF Fellows paper, 2023, 3 pgs.
- Sturtevant, Capt. Chelsey, "AI-HyperCal: In-Scene Hyperspectral Imagery Calibration Using AI Known-Point Identification," SOS AUAR, 2021, 11 pgs.