Caine's Warning Sets the Bar for Autonomous Combat Aircraft
Gen. Dan Caine, chairman of the Joint Chiefs of Staff, told an Air Force conference this week that American troops must now assume they will be hunted by autonomous systems, jammed across the electromagnetic spectrum and tracked in real time. According to DefenseScoop's account of his keynote at the Air and Space Forces Association's Air, Space and Cyber Conference, those three conditions are baseline planning assumptions, not worst-case scenarios. For the teams building autonomous combat aircraft, the speech amounts to a public statement of requirements.
Caine built his case on Ukraine. He described first-person-view drones operating in heavily contested electromagnetic environments, some of them using AI-enabled computer vision to keep driving toward targets after GPS and data links are severed. DefenseScoop reported that the result is affordable precision at the small-unit level, and a pace of change at the tactical edge that software and autonomy now set. Caine cited one figure to make the point: at certain points along the front line, a newly arrived Russian recruit can expect to survive as little as 20 to 30 minutes.
That number matters more than a threat briefing because of the mechanism behind it. Low-cost precision weaponry, fielded quickly, adapted continuously and scaled across a battlefield, is what compresses that timeline. Caine's stated conclusion is that the cost of a lethal exchange keeps falling. He named close air support as the mission he is least certain about, which is also the mission where the cost gap between inexpensive attritable systems and crewed aircraft is hardest to argue away.
Ukraine is the reference case because the experiment has already been run there. FPV drones began as consumer hardware and ended up as a precision strike system working without GPS and without a dependable link, the combination Caine described. The lesson for buyers is that the capability curve is being set by iteration speed, not by the size of a development budget, and that the units closest to the fighting are doing most of the iterating.
Three Conditions, Three Requirements
Caine's formulation is specific. Being hunted, jammed and tracked are three separate problems, and each one maps onto a distinct engineering requirement for the aircraft and software the Pentagon is now buying.
| Stated condition | What it demands |
|---|---|
| Formations hunted by autonomous systems | Autonomous detection and terminal guidance, with no human in the loop |
| Jammed across the spectrum | Navigation and datalinks that survive contested electromagnetic environments |
| Tracked in real time | Lower signatures and tighter emissions discipline across the formation |
Each condition carries a cost that lands on whoever is building the autonomous combat aircraft. Terminal guidance without a human in the loop raises target-verification questions the U.S. military has not settled. Jam-resistant navigation means funding inertial and visual alternatives to GPS. Countering real-time tracking pushes toward lower signatures, which trades against the data-sharing that makes crewed-uncrewed teaming work at all.
Jamming is the condition with the widest blast radius. U.S. formations depend on datalinks to build a shared picture of the battlefield, and a spectrum that cannot be relied on turns a networked advantage into an exposed one. That is why the drone evidence from Ukraine lands so heavily. A weapon that keeps working when the link dies is worth more than a better weapon that stops.
The implication for crewed platforms is uncomfortable. Close air support has rested on a presumption that friendly aircraft can find and reach a target without being attrited on the way in. A battlefield where small units hold affordable precision, and where formations are tracked in real time, erodes that presumption. Value shifts toward systems that can be lost without the mission being lost with them.
Who Builds the Autonomous Combat Aircraft
The threat description is familiar. The keynote matters because of the audience and the timing. According to DefenseScoop, Caine was addressing the service furthest along in fielding uncrewed combat aircraft, and he did so during a conference week when the industrial base for those aircraft was on display in the same halls.
The Air Force is advancing toward pilotless fighters designed to fly alongside crewed jets. The Pentagon's name for the concept is Collaborative Combat Aircraft, or CCA. Production is concentrated. General Atomics and Anduril hold the only two Pentagon production contracts for uncrewed CCA, according to the Washington Times. At the same conference, Lockheed Martin and Kratos Defense showed their own CCA model designs.
| Company | Position on CCA work |
|---|---|
| General Atomics | Holds a Pentagon production contract for uncrewed CCA |
| Anduril | Holds a Pentagon production contract for uncrewed CCA |
| Lockheed Martin | Presented a CCA design at the AFA conference |
| Kratos Defense | Presented a CCA design at the AFA conference |
Caine also tied a recent rescue of an Apache crew to the case for AI in combat, and to his argument that the joint force must modernize if it is to win a future war. Put the warning and the procurement picture side by side and a gap appears: the demand signal is explicit and public, while the supply base is two firms wide.
The Trade-Off in a Two-Vendor Base
Concentration is not automatically a defect. Two production partners give the Air Force a simpler integration path and a single point of accountability for early fielding. The cost of that simplicity is a limit on how many engineering teams can absorb a fast-moving requirement set for autonomous combat aircraft.
The binding constraint may not be airframes. If autonomy, jam resistance and GPS-denied navigation improve on software timelines instead of hardware timelines, the bottleneck shifts from factory throughput to the pace at which a small number of vendors can rewrite their stacks and requalify them. Lockheed and Kratos presenting designs signals appetite and latent capacity, not awarded work, and design work does not shorten a production line.
Cost discipline sits on the other side of the same coin. Caine's central claim is about exchange ratios, and exchange ratios depend on unit price and production rate as much as on capability. A program that fields a small number of exquisite uncrewed aircraft reproduces the problem the keynote described. A program that fields many inexpensive ones addresses it, and that requires a supplier base able to scale output rather than a single integrator able to perfect a design. Incumbents can counter with integration and sustainment records, which matter more as fleets grow.
What to Watch
Four things decide whether the autonomous combat aircraft buildout keeps pace with the requirement set Caine described.
- Whether the Pentagon widens the CCA production contract set beyond General Atomics and Anduril, or holds the current two-vendor structure through the next acquisition milestone.
- Whether the three conditions Caine named are written into formal requirements for navigation, datalinks and signature control, or stay keynote language.
- Whether unit cost becomes an explicit program metric alongside capability, given his emphasis on the falling price of a lethal exchange.
- How quickly the Air Force moves pilotless fighters from design work into squadrons flying beside crewed jets.
The Air Force's direction is already set. Pilotless fighters will fly alongside crewed ones, two companies are contracted to build them, and at least two more are showing designs in the hope of joining. Caine's keynote supplies the doctrinal justification for that path and raises the cost of treating autonomy as a problem for the next decade rather than the current budget cycle.
Why this matters
Caine's warning is best read as a requirement document delivered in public. Once the top U.S. officer states that formations will be hunted, jammed and tracked, autonomy moves from research line item to baseline expectation for every platform the joint force buys. The narrowness of the CCA production base is the distance between that expectation and the industrial capacity behind it.
Photo by Benjamin Chambon on Unsplash
Related Articles
- DHS Counter-Drone Contract Vehicle Reopens Anduril's Product Bid
- Anduril Takes Both Tracks of the $1.5B DHS Counter-UAS Contract
- Pentagon's AI Counter-Drone Marketplace Puts Procurement Speed to the Test
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.