Emerging technologies, such as artificial intelligence, have the potential to disrupt traditional concepts of operation involving nuclear assets, especially bombers. What capabilities and intent do adversaries possess to utilize advanced technologies to hold Air Force Global Strike Command (AFGSC) assets at increased risk?
In the context of our own networks, we must examine the specific opportunities and risks of incorporating artificial intelligence into nuclear command and control (NC2) systems, focusing heavily on maintaining safety, security, and strategic stability. If it is determined that AI should not be integrated into NC2, where else could AI be safely used to support the Nuclear Enterprise?
Finally, as we evaluate the security of these advanced networks, we must also evaluate the overall vulnerabilities and resilience of nuclear command, control, and communication systems to cyberattacks. Ultimately, what is the potential for these cyber vulnerabilities or an adversary cyberattack to escalate to a nuclear conflict?
- Bergin, Capt. Connor T., "Beyond Brinksmanship: How Evolving Nuclear Deterrence Endangers Strategic Stability," AFGC thesis, 2025, 43 pgs.
- Bergin answers this by detailing how disruptive capabilities—such as AI-enabled targeting, persistent ISR, hypersonic weapons, and low-cost drone swarms—threaten the survivability of second-strike nuclear forces like bombers and ballistic missile submarines. He highlights the severe danger of "entanglement," where dual-use early warning and command-and-control systems blur the line between conventional and nuclear operations, vastly compressing decision-making timelines and risking inadvertent escalation. To preserve stability against these technological threats, Bergin recommends augmenting physical hardening with counter-drone and counter-hypersonic defenses, deception tactics, and the development of conventional-only command and control systems separated from nuclear networks.
- Boben, Maj. Joseph J., "Artificial Intelligence and Its Impact on the Low-Observable Advantage," AF Fellows (Oak Ridge), 2025.
- Answers by illustrating how AI algorithms applied to air defense systems can enhance signal processing and constant false alarm rate (CFAR) techniques to extract low-observable (LO) targets, such as the B-2 and B-21 stealth bombers, from heavy clutter and noise. However, Boben concludes that while AI increases the lethality of adversary defenses, it also introduces frailties such as a reliance on vast training data, susceptibility to data poisoning, and algorithmic bias. Ultimately, he answers the prompt by demonstrating that stealth technology will remain a relevant advantage for the bomber fleet if the U.S. strictly manages its physical and virtual signatures and develops robust counter-AI tactics.
- North, Matthew, "Can Machines Have Ethics?" AFGC thesis, 2025, 44 pgs.
- North warns that utilizing autonomous AI/ML in offensive cyberwarfare could be "savage". A human operator can limit the scope of a cyberattack to abide by the rules of war, but an AI instructed simply to "destroy the enemy" will calculate the fastest, most permanent method available. He points out that an unrestricted AI might choose to melt down a nuclear power plant, shut down hospital grids, or crash airliners, causing devastating and indiscriminate collateral damage that no human commander would authorize.
- Schuyler, Maj. Amanda E., "Nuclear Command, Control and Communications Modernization: Making U.S. Nuclear Forces Competitive for Great Power Competition," AF Global College thesis, 2024, 45 pgs.