Air Force Doctrine Note 25-1, *Artificial Intelligence*, presents AI as essential to maintaining U.S. airpower and strategic advantage while insisting that adoption remain ethical, secure, and human-centered. The doctrine argues that AI can process information rapidly, improve intelligence, surveillance, and reconnaissance, support targeting and resource allocation, strengthen logistics and predictive maintenance, and sustain command and control in contested environments. Examples include semi-autonomous airlift delivering more than 20 urgent parts orders, AI-enabled ISR detecting pop-up threats through multimodal data fusion, resilient communications that reroute around failed nodes, Autonomous Collaborative Aircraft enhancing crewed platforms, and PANDA predicting component failures to increase aircraft availability.
A central distinction is between automation, which performs narrow repetitive tasks, and autonomy, which pursues multiple goals under changing conditions. Both must operate within defined authorities, with human supervision calibrated to risk. The doctrine therefore favors human-machine teaming rather than replacing decision-makers. Human judgment becomes especially important in high-consequence missions, including nuclear deterrence, where context, ethics, wisdom, and awareness of broader consequences exceed current AI capabilities.
The document also emphasizes that AI performance depends on trustworthy data, secure infrastructure, effective governance, experimentation, and an AI-fluent workforce. Bias, incomplete or poisoned data, adversarial manipulation, cyberattack, model inversion, deception, and edge-case failures require diverse and verified datasets, explainability, audits, testing, access controls, encryption, intrusion detection, and continuous human oversight. Experimentation and training should build confidence without allowing expertise to atrophy.
Finally, AI adoption is framed as a necessity in competition with China and Russia and in contested communications environments supporting ABMS and JADC2. Leaders should choose AI only when its advantages in speed, scale, or prediction outweigh costs, vulnerabilities, complexity, and the risks of replacing simpler, more reliable methods.
Article timeline
Build this article’s timeline
Extract dated events, key actors, and open questions using the same timeline analysis available in Signal.
Key Phrases:Artificial Intelligence (AI)100%Air Force Doctrine Note 25-198%human-machine teaming95%automation versus autonomy91%responsible AI94%AI-enabled capabilities90%machine learning89%big data analytics84%Intelligence, Surveillance, and Reconnaissance88%autonomous agents and intelligent swarms86%
Comments
Login to view and post comments.