King’s Wargaming Week gathers serving officers, defense academics and industry practitioners. This year’s program blended on-stage panel talks with hands-on tabletop sessions, showcasing new designs and debating what should come next.
If you’ve never attended, picture twenty professionals, role-playing heads of state. In one session, I haggled over Filipino air-base access while an academic played XI Jinping, offering “in-kind” fighter-jet-joy-rides. Decisions were whispered to adjudicators, who slipped behind a curtain to roll dice and consult matrices. It’s tactile, human, brilliant—and achingly slow.
Look up from the map and you notice what isn’t there: no live dashboards, no automated logs, not even basic speech-to-text. In King’s new NATO game Horizons, game-play paused every ten minutes so a facilitator could re-type our moves into Excel. In 2025. Then came the zinger from a speaker: “AI is practically useless in wargaming.”. The fear is understandable: if AI can’t explain itself, how can we trust it to influence games that should inform real policy? Fair question—but solvable.
The common, but flawed, reflex is to assume the only way to use AI is to hand it the reins—treating it as an equal strategist and sidelining human expertise. In reality, AI is a tool that thrives under clear guard-rails. That’s the faster-horse mentality. AI isn’t a shiny saddle; it’s the engine room. Treating it as a gimmick wastes the opportunity—and the obligation—to modernise how we train and, ultimately how we fight.
At Hadean, we follow one principle: Immediately Validatable Output (IVO). AI should only generate content a human can check on the spot.
- Transcripts of what was just said? Play back the audio.
- Synthetic social-media blasts? Let a subject-matter expert confirm tone and plausibility.
- Narrative twists tied to dice rolls? Compare to doctrine and probability tables. If you can’t validate it now, don’t use AI for that task. Accountability stays intact, speed goes up.
So, how can we harness AI without throwing the baby out with the bathwater?
- Keep the mechanics, streamline the admin. Computer vision can read token states; cloud speech services can live-transcribe dialogue; and AI can add narrative flair to dice rolls—each step IVO‑checked so players spend time deciding, not databasing.
- Bridge the table to full-spectrum simulation. Keep players pushing counters and negotiating face-to-face, while AI pipes their moves into a multi-domain model (land, sea, air, space—plus political, economic, social, technological, legal and environmental factors). Results come back instantly, letting humans challenge outcomes without breaking immersion.
- Embrace the derailment. The dirty reality is we can’t cling to today’s format forever. One panel echoed the Marlinspike thesis, an argument that decision cycles are collapsing and human‑only play can’t keep up. The next step may not look like wargaming as we know it: the same synthetic environments and wargame rulesets we use for play must morph into live decision‑support tools, letting AI partner with commanders at machine speed—and that leap will save lives
Take Populating the World of Training as an example – a project we delivered with British Army Futures and DSTL that explored how AI can support military training exercises. Trainees still strap on their TES vests and run their lanes, achieving the same training outcomes they’re used to; AI fills the battle-space with civilians, traffic and social chatter, then harvests every round and radio call for rapid, data‑rich after‑action review.
I enjoyed Wargaming Week—as always—the games were lively and the debates honest. Yet, even under a banner of gaming the future, we tip‑toed around the real digital revolution, let alone the AI revolution. Until we treat technology as integral rather than ornamental, we’ll still be rehearsing today’s fights at yesterday’s tempo.
King’s Wargaming Week traditionally kicks off the summer wargaming circuit, with DSET next month and Connections UK in September keeping the momentum alive. Get in touch to meet us there and discuss how AI‑enabled workflows can slash admin, deepen analysis and bridge the gap between training and operations.