General Sir Roly Walker's admission at RUSI this week that Project ASGARD has collapsed corps-level planning cycles from 72 hours to one isn't a promise about the future. It's a public accounting of something already deployed.
Speaking at the RUSI Land Warfare Conference on June 23, Walker laid out numbers that would have sounded implausible three years ago. Where a corps headquarters once needed three days to complete a planning cycle, it now needs one hour. Where it could prosecute 24 targets a day, it can now do ten times that. The systems producing those results aren't prototypes. They're running. And the vendors behind them, Anduril and Helsing, are no longer defense-tech hopefuls quietly demoing in sandboxes. They're proven infrastructure inside a NATO army, and their valuations show it.
Project ASGARD was announced in October 2024, contracts awarded in January 2025, and a prototype deployed just four months later for NATO's Exercise Hedgehog in Estonia. That speed would be remarkable for any defense procurement. For a UK Ministry of Defence program, it's essentially science fiction. ASGARD connects Anduril's Lattice mesh network, which routes sensor-to-shooter data across the battlespace in real time, with Helsing's Altra edge-AI software, which runs on drones to identify and fingerprint targets from full-motion video and pass targeting data to third-party systems. The command layer runs on Sitaware C2 software. The combined effect is a corps headquarters that no longer needs to manually collate intelligence, convene planning cells, and work through targeting bureaucracy across multiple days. The machine does the aggregation. Commanders say yes or no.
Walker called ASGARD "a digital juggernaut that is evolving every 8 to 12 weeks" and explicitly framed it as the foundation of an agentic AI headquarters, not merely a faster version of existing command and control. That framing matters. It means the British Army isn't just digitizing old processes. It's building toward a headquarters where AI runs the workflow and humans approve outcomes, which is a genuinely different relationship between commander and machine than anything that existed before Ukraine changed what modern land warfare looks like.
Walker's goals are explicit: double lethality by 2027, triple it by 2030, and eventually reach a tenfold increase. The 20/40/40 force-mix underpins that ambition, with 20% of capability in traditional platforms like tanks, 40% in expendable autonomous systems, and 40% in reusable AI-enabled assets. That framework emerged from the 2025 Strategic Defence Review, and ASGARD is the clearest operational proof that the model works. The Army already has 10,000 small drones delivered, more than £100 million committed to land drone swarms, and Project AXA running as a fast-track procurement channel for attritable systems.
For Anduril and Helsing, the significance of this week's RUSI announcement isn't only reputational. Walker just provided the most credible publicly verifiable proof point either company could ask for: a named NATO general, at a named conference, citing specific metrics from a deployed system. In defense procurement, that kind of reference is worth more than any demo. Allied militaries shopping for AI-enabled C2 and targeting capability now have a live benchmark and two vendors directly associated with it. The market has already priced some of this in, but probably not all of it. Anduril closed a $5 billion raise in May 2026 at a $61 billion valuation, roughly double where it sat eleven months earlier following a $2.5 billion Founders Fund-led round at $30.5 billion, according to reporting by Augment Market. Helsing's own $1.2 billion raise, led by Dragoneer and Lightspeed Ventures and reported in May 2026, put the Munich-based company at an $18 billion valuation, nearly 30% above the €12 billion it commanded after Daniel Ek's €600 million Series D in June 2025. Those rounds preceded the RUSI announcement. Both companies now have a live operational credential they didn't have before Walker took the stage.
Helsing's recent contract pipeline makes the direction clear. In February 2026, the Bundeswehr awarded the company an initial €269 million deal for HX-2 loitering munitions, with a framework option reaching €1.46 billion. Anduril, meanwhile, secured a $20 billion Army IDIQ vehicle in March 2026, the largest ever awarded to a non-traditional contractor by the US Army, alongside a Dutch Ministry of Defence counter-UAS contract announced in May. Neither company is waiting for NATO allies to catch up. They're running contracts in parallel across multiple governments.
Frankly, what Walker described changes the procurement calculus for any allied army that watched that keynote. The question is no longer whether AI can deliver measurable operational impact at the corps level. The British Army answered that. The question now is how fast everyone else can replicate it, which vendors are already embedded in the right allied bureaucracies, and whether the next ASGARD equivalent goes to Anduril and Helsing by default or whether European and American competitors get a genuine look. Given both companies' existing NATO footprint and the 8-to-12-week iteration cycle Walker cited, the head start is real and it's growing.
Also read: A San Jose legal-tech startup just sued the US government over losing Anthropic's Fable 5 and it won't be the last • SK Hynix is betting $29 billion that the AI memory boom is nowhere near over • ASE Technology raises its 2026 capex to $8.5 billion as AI chip packaging demand overwhelms supply