Washington, D.C. — The Syracuse University Institute for Artificial Intelligence (SIAI) will convene a two-day workshop this fall on a question that has become hard to avoid in defense research: what happens to AI-driven decision making when an adversary is actively working to jam, deceive, or sever the systems it depends on.
The Workshop on Contested and Distributed Multi-Domain Decision Making: AI Driven Action across EW, Radar, Aerial, and Space is sponsored by the Army Research Office and runs Oct. 12-13, 2026. SIAI is holding the event at the Syracuse University Washington, D.C. Center (SUDC), which places the discussion a short walk from many of the agencies with a direct stake in the outcome.
The workshop is invitation-only. Instead of an open call for papers, the organizers assembled a roster of researchers already working these problems, with the aim of keeping the room small enough for genuine technical argument rather than a parade of conference talks. Researchers interested in future activity can ask to be added to the interest list by contacting the workshop's principal investigator.
The underlying problem is easy to state and hard to solve. Multi-domain operations increasingly lean on machine learning to make sense of sensor data and to act on it across electronic warfare, radar, aerial, and space domains. Adversaries know this. Electronic attack techniques such as jamming degrade the sensor picture and, with it, the navigation and planning capability of the platforms that rely on it. Teams of humans and autonomous agents need to hold a coherent shared picture of the world even when the bandwidth to synchronize it is limited, intermittent, or under attack. Several of the workshop's stated objectives follow from that: positioning EW assets to disrupt an opponent's own AI-enabled decision making, keeping autonomous behavior stable under interference, managing spectrum allocation under adversarial pressure, coping with distribution shift caused by interference patterns, and tracking maritime vessels that have switched off their automatic identification systems.
Two keynotes are confirmed. DARPA Program Manager Eric Davis opens the first morning, and Laurence Mixon, acting deputy capability program executive for the Army's Intelligence and Spectrum Warfare portfolio (CPE ISW) at Aberdeen Proving Ground, opens the second. Each keynote runs forty minutes with a ten-minute question period.
The rest of the draft agenda is a deliberate mix. On the government and federally funded side, the program includes speakers from the Institute for Defense Analyses, the Naval Information Warfare Center, and the Army Cyber Institute at West Point. Industry participation spans large primes and smaller specialized firms: Lockheed Martin, Epoch Geo, Hidden Level, MyDefence, Nexcepta, and Boston Fusion. The academic contingent includes the Syracuse, UT Austin, Virginia Tech, Cornell, the University of Virginia, and the Institute for Human and Machine Cognition Roughly two dozen speakers are on the schedule across the two days.
The format is built to encourage exchange. Talks are grouped in clusters of three, and each cluster is followed by a fifteen-minute panel with those speakers taking questions together. That structure is repeated through both days, with the second afternoon reserved for a longer closing discussion on next steps. The workshop is organized by two Syracuse Institute for AI faculty members: Paulo Shakarian and M. Cenk Gursoy.