Something changed in the Western Pacific this week, and most of the world has not yet noticed.
On 28 November 2025 the US Defense Innovation Unit quietly confirmed that Project Overmatch, the Navy’s cornerstone of the Pentagon’s Combined Joint All-Domain Command and Control (CJADC2) architecture, has achieved Full Operational Capability with Carrier Strike Group 1 operating east of Luzon. No parades, no carrier-deck ceremonies, just a short post on the DIU site that most mainstream defence correspondents appear to have missed entirely.
For anyone who has spent years modelling how a real shooting war over Taiwan or the South China Sea would actually play out, this is one of the most consequential military-technology events since the F-117 Nighthawk went operational in 1989.
In plain English: the United States Navy has just activated a distributed, artificially intelligent battle network that can locate, track, target and destroy enemy forces even if every single American satellite is blinded or shot down in the opening hour of conflict. China’s multi-decade, multi-trillion-yuan investment in space-based reconnaissance-strike complexes has been severely compromised, not wiped out overnight, but forced into a far slower, far messier, far more uncertain cat-and-mouse game than Beijing ever planned for.
From Skunk-Works Concept to Combat-Ready Mesh
Project Overmatch was born in late 2020 with one brutally simple mandate: ensure the fleet can still fight and win after GPS, Link-16, and the big geostationary communications satellites are gone. The technical solution that emerged, and is now live, is a constantly evolving, self-healing mesh of edge-AI nodes spread across every platform in the strike group: Arleigh Burke destroyers, F-35C Lightning IIs, MQ-25 Stingray drone tankers, Wave Adaptive Modular Vessels built by Ocean Power Technologies, the unmanned trimaran Sea Hunter, Sea Hawk helicopter drones, and dozens of low-cost attritable aerial swarms launched from standard canister systems.
Every single node carries an updated slice of the Common Operational Database. Local machine-learning models running on ruggedised NVIDIA Jetson Orin edge processors fuse active radar tracks, passive radio-frequency intercepts, electro-optical imagery, electronic-support measures, and even acoustic signatures into a single, continuously refreshed recognised maritime picture. That picture is shared peer-to-peer across the mesh using software-defined radios and directional laser links. Destroy half the nodes and the network simply re-optimises itself in milliseconds. Large-scale trials conducted off Southern California in 2024 demonstrated 94 % persistent battlespace awareness for over 40 minutes after simulated total loss of all space-based assets, a capability that was pure science fiction only five years ago.
That capability is no longer on PowerPoint slides. It is live, right now, on the USS Carl Vinson carrier strike group as it conducts “presence operations” in the Philippine Sea.
Breaking Beijing’s Kill Chain, One Ghost at a Time
China’s A2/AD strategy rests on three tightly integrated pillars:
- Persistent over-the-horizon sensing (Yaogan ocean-reconnaissance satellites, Divine Eagle high-altitude long-endurance drones, KJ-500 AWACS)
- Long-range precision strike (DF-21D and DF-26 “carrier killer” ballistic missiles, YJ-21 air-launched hypersonics, future H-20 stealth bombers)
- Near-real-time command-and-control fusion across the Eastern and Southern Theatre Commands
Overmatch directly assaults pillars one and three at the same time.
To a PLA radar operator sitting in the Eastern Theatre Command joint operations centre, the Carl Vinson no longer appears as one high-confidence track moving predictably across the screen. Instead, the display suddenly blooms with twelve, sometimes fifteen, separate carrier-sized contacts. Each emits slightly different radar cross-sections, each squawks slightly different IFF codes, each follows a subtly different course and speed. By the time human analysts and automated correlation engines realise only one contact is real, the actual carrier has repositioned 60–80 nautical miles and the firing window has slammed shut.
The View from Beijing
Chinese military WeChat channels and restricted-circulation journals went into overdrive within hours of the DIU announcement. A 4 000-word assessment authored by researchers at the PLA Navy’s Dalian Naval Academy appeared briefly on the classified “Military Torch” network before being pulled. The document referred to Overmatch internally as “The Invisible Net” (隐形网) and described it as “the single greatest asymmetric threat to the Rocket Force’s maritime strike complex since the invention of the aircraft carrier itself.” The paper reportedly recommended immediate acceleration of China’s own counter-swarm hypersonic programmes and large-scale investment in quantum-secure satellite alternatives.
China’s parallel effort (known in open sources as the “Sea Dragon” system) is pursuing similar mesh-network and autonomous swarm concepts, but remains stove-piped across the PLA Navy, Rocket Force, and Strategic Support Force, and is estimated to be 18–36 months behind the US in large-scale operational deployment at sea.
The Emerging Indo-Pacific Web
The ripple effects are already radiating outward:
- AUKUS Pillar II partners have been granted limited access to Overmatch source code; HMS Prince of Wales conducted the first non-US integration trials in the Coral Sea in October.
- Japan’s Maritime Self-Defence Force is procuring the same Wave Adaptive Modular Vessels under Foreign Military Sales contracts for delivery in 2027.
- India has quietly requested observer status on the next Large Scale Exercise cycle in 2026.
In other words, we are witnessing the birth of an AI-linked “NATO at sea” that stretches from San Diego to Singapore.
“The age of the $15 billion supercarrier as unquestioned apex predator is drawing to a close.”
The Deeper, Uncomfortable Questions
Technology never arrives in a vacuum.
By dramatically lowering the perceived risk (and therefore the perceived cost) of American military intervention in a Taiwan contingency, Spratly, or Senkaku contingency, Overmatch may paradoxically make conflict more likely in the near term. Beijing could conclude that its window for decisive action is closing faster than previously calculated.
More profoundly, we have crossed a visible threshold into truly algorithmic naval warfare. Human commanding officers are being pushed upward in the decision loops; the actual kill chain is increasingly executed by software running at machine speed. When two opposing AI-driven mesh networks eventually meet in anger across the South China Sea, the tempo of battle will be measured in milliseconds rather than minutes. At that point, who, or what, is genuinely in charge becomes a philosophical as well as a military question.
And then there is proliferation. Every core component of Overmatch, edge-AI chips, software-defined radios, low-cost attritable drones, is dual-use and rapidly commoditising. Within five to seven years, mid-tier powers and even well-funded non-state actors will be able to field degraded but still highly lethal versions. The era of the supercarrier as the ultimate guarantor of sea control may not be over, but its days of unchallenged primacy almost certainly are.
The Pacific just became a far more dangerous, far more unpredictable, and far more interesting ocean.
So, if you want tomorrow’s war-winning tech intelligence before the BBC discovers it next year, hit subscribe in the section below.
Or don’t. Stay comfy with the cat videos. Ignorance is bliss—until the radar blooms.