Somewhere over a restricted test range, an autonomous combat drone is reportedly flying missions without a pilot on board or a hand on the stick. It may be joined by a second. And if the U.S. Air Force’s plans hold, both will eventually fly alongside a human in the cockpit of the most advanced fighter jet the country has ever built.
On March 21, 2025, President Trump stood with Defense Secretary Pete Hegseth and formally designated the F-47 as the crewed centerpiece of the Next-Generation Air Dominance program, or NGAD. The announcement confirmed what defense insiders had anticipated for months: the Air Force is building a sixth-generation fighter designed from the start to operate not alone, but as the command node for a pack of AI-driven drone wingmen known as Collaborative Combat Aircraft, or CCAs.
Two of those drones, reportedly called “Dark Merlin” and “Fury,” have been described in defense media accounts as already undergoing flight testing. If accurate, the Air Force is further along in fielding autonomous combat partners than most of the public realizes.
The F-47 and the system around it
The F-47 is not just a new airplane. Official Defense Department descriptions frame NGAD as a networked “family of systems” in which the crewed jet serves as the brain and its drone wingmen act as extended limbs. A single F-47 pilot could direct several CCAs during a mission, multiplying the squadron’s sensor coverage, weapons capacity, and survivability without a matching increase in trained aviators or costly manned airframes.
Boeing was widely reported as the prime contractor selected to build the F-47 following the March 2025 announcement. Air Force Chief of Staff Gen. David Allvin spoke publicly about the effort, framing it around human-machine teaming at scale. A Congressional Research Service analysis (IF12805) independently confirmed the program’s formal status, its funding trajectory, and ongoing congressional oversight.
Under the concept of operations the Air Force has outlined, the F-47 would carry the most advanced radar, electronic warfare suites, and battle management software. The CCAs, meanwhile, could be tailored for specialized roles: flying ahead as decoys to draw out enemy air defenses, carrying additional weapons for strike missions, or jamming adversary radar from positions too dangerous to risk a human pilot. By distributing risk across cheaper, expendable airframes, commanders could push drones into the most contested airspace while keeping the F-47 and its pilot at a safer distance.
That approach directly addresses a math problem the Air Force has struggled with for years. The service’s fighter fleet has been shrinking as legacy jets age out faster than replacements arrive, and the cost of manned fighters keeps climbing. If CCAs can be produced at a fraction of the cost of a crewed jet, a mixed squadron of one F-47 and several drones could deliver more combat power per dollar than a formation of manned aircraft alone.
Dark Merlin, Fury, and what we actually know
The names “Dark Merlin” and “Fury” have appeared in reporting from outlets including The War Zone and Aviation Week, which have tracked the CCA program closely. In 2024, the Air Force publicly awarded Increment 1 CCA contracts to Anduril Industries and General Atomics, confirming that at least two companies were building drone prototypes intended to fly alongside next-generation fighters. Anduril’s entry has been linked to the “Fury” name, while details around “Dark Merlin” remain less clearly attributed.
What has not happened is official Pentagon confirmation tying those specific names to flight-test milestones. Neither the March 2025 DoD announcement nor the CRS analysis mentions Dark Merlin or Fury by name, and no public Air Force record available as of June 2026 independently verifies their testing status. Defense programs at this classification level routinely operate behind restricted information controls, and flight-test activity at classified ranges would not necessarily surface in public documents.
That gap matters because the timeline for integrating autonomous wingmen with the F-47 is one of the program’s biggest open questions. The Air Force has discussed initial operational capability targets in the early 2030s, a timeline that leaves relatively little margin for the kind of software integration, autonomy certification, and tactics development that a manned-unmanned team will require. Whether Dark Merlin and Fury are flying today or still working through ground testing has real implications for whether that schedule is achievable.
The cost question no one has answered publicly
Sixth-generation fighters push the boundaries of stealth, sensor fusion, and propulsion technology, and they carry price tags to match. The F-35, a fifth-generation jet, costs roughly $80 million per copy depending on the variant. The F-47 is widely expected to exceed that figure, though no official unit cost has been released.
The economic argument for CCAs rests on volume. Former Air Force Secretary Frank Kendall discussed a target of roughly 1,000 CCAs to complement the crewed fleet, with unit costs he hoped to keep in the range of $20 million to $30 million, far below a manned fighter. If production scales as planned, the blended cost of an NGAD squadron could come in below what an all-manned force would require. But those figures remain projections. No confirmed per-unit price for the F-47 or any named CCA variant has appeared in verified budget documents, and defense procurement history is littered with programs whose costs ballooned well past early estimates.
Autonomy, ethics, and the rules of the fight
How much independence these drones will have in combat is another unresolved question. Public Air Force descriptions emphasize “collaborative” behavior and human oversight, but they do not spell out the balance between pre-programmed flight plans, onboard AI decision-making, and real-time pilot commands. That balance has consequences for rules of engagement, software certification, and cybersecurity.
The Pentagon’s existing policy on autonomous weapons requires a human to authorize the use of lethal force, but the practical mechanics of applying that standard to a fast-moving air battle with multiple drones are still being worked out. A pilot managing three or four CCAs simultaneously may not have time to approve every individual action, which means the AI will need enough authority to maneuver, defend itself, and execute pre-approved tasks without constant human input. Where exactly that line falls will shape how the Air Force certifies these systems and how allies and adversaries perceive them.
Why the strategic stakes are so high
The urgency behind NGAD is not abstract. China has been developing its own sixth-generation fighter program and has publicly demonstrated loyal wingman drone concepts, including the FH-97 and GJ-11 platforms shown at recent air shows. The People’s Liberation Army Air Force is investing heavily in AI-enabled combat systems, and U.S. defense officials have repeatedly cited the pacing threat from China as the primary driver behind NGAD’s timeline.
If the Air Force delivers a functioning F-47 integrated with capable, affordable autonomous wingmen, it will have redefined what a fighter squadron looks like and how air campaigns are planned. A distributed force of crewed and uncrewed aircraft would be harder for an adversary to target, more resilient to losses, and able to cover more airspace than a traditional formation of manned jets.
If, on the other hand, technical hurdles, cost growth, or shifting political priorities slow the effort, NGAD could join a long list of ambitious aerospace programs that promised transformation but delivered only incremental change. The B-21 Raider bomber, which recently entered flight testing after years of disciplined cost management, offers one model for how a classified program can stay on track. The Army’s Future Combat Systems, canceled in 2009 after billions in spending, offers the cautionary counterexample.
What to watch as the program moves forward
The confirmed facts are substantial: the F-47 is real, funded, and publicly endorsed at the highest levels of government as the centerpiece of a new air dominance architecture. The CCA effort is documented in congressional budget lines and backed by contracts with at least two major defense firms. What remains unconfirmed are the precise identities, capabilities, and readiness levels of individual drone designs said to be flying today.
For readers tracking this program, the signals to watch are straightforward. Official acknowledgment of specific CCA flight-test milestones would move Dark Merlin and Fury from informed speculation into confirmed fact. A public flight demonstration of manned-unmanned teaming would show the concept works outside a simulation. And the fiscal year 2027 budget request, expected in early 2026, will reveal whether Congress is funding NGAD at the levels the Air Force says it needs or forcing the service to make tradeoffs that could stretch the timeline.
Until those markers appear, the story of the F-47 and its drone wingmen sits at a pivotal but incomplete stage: ambitious enough to reshape American air power for a generation, and uncertain enough that the hardest questions remain behind classified doors.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
