STRATCOM Symposium Treats AI as a Present Challenge for Nuclear Command

AI moves from future concern to current issue

In early August, United States Strategic Command (STRATCOM) hosted its annual Deterrence Symposium in Omaha under the theme "Peace Through Strength: Turning Defense Strategy into Decisive Deterrence Action." Speakers included the Vice Chairman of the Joint Chiefs of Staff, General Christopher Mahoney, STRATCOM Commander Admiral Richard Correll, and the Commander of U.S. Special Operations Command, Admiral Frank Bradley.

IST's Nuclear Policy team attended to engage practitioners and discuss emerging findings from its work on integrating AI into nuclear command, control, and communications (NC3). The team met officials from several U.S. government agencies, including STRATCOM and the Department of Energy.

According to Cortino, panelists described AI as something that can produce countless new operational advantages but also introduces myriad risks. At earlier symposiums, emerging and disruptive technologies were typically described as a challenge for the next generation. This year, IST saw concrete discussion of AI as a present threat vector that must be addressed and mitigated immediately. Admiral Correll also spoke about using AI for operational advantage through faster response times.

Deterrence by denial and the NC3 "system of systems"

A second takeaway was the growing role of deterrence by denial over deterrence by punishment. Both terms originated in a 1959 monograph by Glenn Snyder. Denial works by convincing an adversary that its efforts will fail, for example that a surprise attack will be stopped. Punishment seeks to convince an adversary that a surprise attack would meet a severe response.

IST notes that antiballistic missile systems such as the Golden Dome initiative operate on deterrence by denial and are controversial, because they could introduce doubt about whether destruction would truly be mutually assured. Cortino also points to the National Security Strategy's directive to build U.S. military infrastructure in the First Island Chain to deny any attempt to seize Taiwan. He describes this as a very explicitly clear example of denial logic in a public-facing document.

The third takeaway concerned the structure of NC3. The symposium surfaced the idea that the enterprise comprises more than 200 subsystems, such as infrared satellites that detect the heat signature of a missile launch. According to IST, this means AI integration cannot happen within a single monolithic system. It will take place across a large, heterogeneous architecture with different missions, requirements, and vulnerabilities.

IST's AI-NC3 project focuses on three subsystems in particular: early warning, adaptive targeting, and decision support.

Broader agenda and site visit

Nearly every panel included discussion of AI, according to IST. Its AI-NC3 research drew particular attention when a panel moderator referenced discussions from an IST workshop held in San Francisco in April 2026. That workshop focused on Dr. Kenneth Payne's paper on AI wargaming and nuclear escalation.

The symposium also covered competition in the electromagnetic spectrum, constraints facing the defense industrial base, and greater integration and burden sharing with U.S. allies and partners.

On the third day, the team visited Offutt Air Force Base, STRATCOM's headquarters. The visit included static aircraft displays, the Defenders of Freedom Air & Space Show, a tour of the E-4B National Airborne Operations Center, or "Doomsday Plane," and a look at the E-6B Mercury, which supports strategic communications and can provide airborne command and control for U.S. nuclear forces.

What it means for the AI-NC3 work

IST concludes that questions around AI and NC3 are moving rapidly from speculative debates toward practical questions of integration, modernization, and strategic stability. The discussions in Omaha will inform the AI-NC3 working groups as they develop final research and recommendations. These will then be used to advise offices across the nuclear enterprise on optimally safe integration pathways.