Drones, not jets: how the next air war rewrites the procurement script
A Turkish drone-maker says the era of the manned fighter is closing. Western air forces should stop treating that claim as marketing.
The most consequential defence prediction of 2026 did not come from a Pentagon briefing room, the Royal United Services Institute, or the Platform One office. It came from Baykar's Selçuk Bayraktar, speaking in the cadence of an engineer rather than a salesman: roughly fifteen thousand fighter aircraft worldwide will, in his telling, "probably" be replaced by unmanned combat platforms. On the same day, he said those platforms are not the next generation of fighter so much as "a different species."
The claim is large, made by a man with standing to make it. Bayraktar's family firm builds the Bayraktar TB2 and the heavier Akıncı, the drones that did for unmanned strike aviation what the iPhone did for touchscreen computing — not by being first, but by being usable. The reception of that claim across NATO procurement offices is the real story, because it forecloses a decade of investment that the West has already pencilled in.
What he actually said
In remarks circulated on 2 July 2026, Bayraktar placed the global manned-fighter inventory at "approximately 15,000" aircraft and argued that the cohort will "probably" be replaced by unmanned systems. He rejected the generational taxonomy that runs F-22, F-35, sixth-gen, seventh-gen, calling AI-powered combat drones "a different species." The frame matters. Generational taxonomy creates a procurement glide-path: a fifth-gen fleet cycles into a sixth-gen fleet on a thirty-year replacement curve, and major air forces stay inside their own sandbox of cost-plus contracts, classified avionics, and export-controlled engines. "Different species" breaks the curve. It implies substitution rather than succession.
The procurement contradiction
Western air staffs are not unaware. The US Air Force has flown Collaborative Combat Aircraft prototypes alongside the F-35 and B-21 programmes. The UK's GCAP / Tempest project has, on paper, an unmanned wingman variant. France's SCAF work with Berlin includes remotely crewed elements. The problem is the budget structure: a sixth-generation manned fighter still absorbs the bulk of the line item, with autonomy layered on as a feature.
Bayraktar's argument, taken seriously, inverts that. If the dominant combat platform of 2035 is unmanned, then the industrial base that needs to scale is software-stacked, batch-producible, and export-frictionless — closer in profile to a Baykar production line in Keşan than to a Fort Worth final-assembly hangar. That is awkward for incumbents, because it depreciates the assets they are currently buying. A $80m F-35 is a sunk cost against a $5m Bayraktar-style combat drone only if the drone cannot survive in contested airspace — and the Ukrainian operational record, however one weights the early-TB2 losses, suggests that with attritable mass the math bends.
The counter-case is straightforward: attritable mass solves low-end fights, not peer air combat. A drone that costs $5m is not, on present evidence, going head-to-head with a Su-57 or a J-20 in contested skies. The honest reading is that both propositions are true at once, and procurement bureaucracies are not built to manage two truths at once. They are built to fund the line item in front of them.
Why Ankara says it louder than Washington
Baykar's industrial incentive is to be the supplier of the new platform, not the buyer. A Turkish firm maximises its addressable market by making the case that the airframe is no longer the strategic object — the software stack, the datalinks, the autonomy pipeline, and the production rate are. None of those assets are protected by the same export controls that govern F-35 subassemblies. That is why a Turkish executive says plainly what a US programme executive cannot: the West's most expensive line items are approaching end-of-life in a domain that no longer requires them.
The structural reading is that defence industrial bases are about to undergo the same platform-to-pipeline displacement that consumer software went through in the 2010s. Capture the integration layer, and the marginal unit cost collapses; lose it, and you keep building bespoke hardware for shrinking export markets. Ankara has positioned itself upstream of the integration layer. The pentagonic incumbents have not.
Stakes, and what remains contested
If Bayraktar is even half right, the procurement consequences land in three places. First, fifth-gen fleet extension programmes become harder to justify in cabinet submissions. Second, export-control regimes built around airframe ITAR / dual-use categories get re-stitched around autonomy software and training data. Third, allied industrial participation in sixth-gen programmes — Japan in GCAP, Italy and the UK likewise — re-prices downward relative to the total programme value, because the airframe share of value is shrinking.
What remains genuinely contested is whether AI-piloted combat aircraft will clear the regulatory and ethical bars in NATO airspace on a timeline that affects 2030s planning. The sources circulated here do not address that question; they assert a market-substitution prediction, not a certification timeline. That distinction matters for finance ministries even if it does not matter for boardrooms. Until that gap is closed, the prediction is best read as an industrial strategy brief from a vendor with a clear interest — compelling, asymmetric, but not yet a procurement fact.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://t.me/ClashReport/94523
- https://t.me/ClashReport/94522
- https://t.me/ClashReport/94521