AI-Capable F-16s: DARPA’s VENOM Autonomy Takes Flight

The idea of an AI piloting a combat fighter jet used to live firmly in the realm of science fiction. Not anymore. DARPA and the U.S. Air Force have crossed a threshold that redefines what it means to fly into battle — and the machine doing the flying is one of America’s most iconic warplanes, the F-16 Fighting Falcon. The program making this possible is called VENOM, and it represents one of the most significant leaps in military aviation since the introduction of fly-by-wire systems.

AI-capable F-16s aren’t just a flashy demonstration of what technology can do. They’re a deliberate, structured step toward a future where autonomous aircraft operate alongside human pilots in contested airspace — faster, more precise, and willing to go where no human should have to. Understanding what VENOM is, how it works, and where it leads gives you a front-row seat to a transformation in air combat that will shape military doctrine for decades to come.

What Is DARPA’s VENOM Program?

Ai-enhanced f-16 fighter jet flying, showing glowing neural network patterns
The f-16, now equipped with darpa’s venom autonomy, takes to the skies with advanced ai capabilities.

VENOM stands for Viper Experimentation and Next-gen Operations Model — Autonomy Flying Testbed (VENOM-AFT). The “Viper” is a beloved nickname for the F-16 Fighting Falcon, and the program’s full name makes its intent unmistakably clear: this is about experimenting with next-generation AI operations on a real, combat-capable fighter.

At its core, VENOM is designed to rapidly test and transition AI autonomy capabilities into uncrewed or optionally crewed combat aircraft. The program falls under the umbrella of DARPA’s broader Air Combat Evolution (ACE) initiative, which aims to build trusted autonomy for air combat by systematically demonstrating that AI can make reliable, split-second decisions in complex aerial environments.

The VENOM Autonomy Kit

The hardware heart of the program is the VENOM Autonomy Kit — an integrated package of hardware and software that gives the F-16 its artificial “brain.” This kit directly interfaces with the aircraft’s flight control surfaces and thrust management systems, allowing an AI to physically fly the jet without human hands on the controls.

The kit isn’t just a bolt-on gadget. It includes a new power distribution system and mission computers integrated into the F-16D’s existing architecture. Critically, it’s designed for rapid AI model updates — engineers can push new algorithms, test them in simulation, then validate them in actual flight with a speed that would have been unthinkable under traditional defense acquisition timelines.

The AI-Capable F-16: Six Vipers, One Mission

Pilot in f-16 cockpit interacting with holographic display showing ai-controlled flight data
Test pilots maintain ‘human-on-the-loop’ oversight, supervising the autonomous f-16’s ai-driven maneuvers.

The Air Force selected the F-16D — the two-seat variant of the Fighting Falcon — as the platform for VENOM testing. The second seat is essential: it’s where the safety pilot sits, monitoring every move the AI makes and ready to take control at any moment. The F-16 wasn’t chosen arbitrarily. It’s a mature, deeply understood platform with decades of operational data, making it an ideal testbed for validating AI behavior in ways that are measurable and meaningful.

Six F-16Ds are being converted for this program. All modifications and testing activities are centered at Eglin Air Force Base, Florida, with the 96th Test Wing, the 40th Flight Test Squadron, and the 85th Test and Evaluation Squadron all playing active roles. Lockheed Martin, the F-16’s original manufacturer, is involved in the modification process.

A Methodical Path to Autonomous Flight

VENOM’s testing didn’t jump straight to letting an AI barrel-roll a fighter jet unsupervised. The program followed a deliberate progression:

Aircraft modification phase: Hardware integration, systems checks, and ground validation.
Simulation phase: AI models trained and stress-tested in high-fidelity virtual environments before touching an actual aircraft.
Piloted flights with autonomy kit engaged: Beginning in late 2023, test pilots flew the modified F-16Ds with the autonomy kit active, evaluating how the AI’s inputs interacted with the aircraft’s existing systems.
In-air autonomy testing: Beginning in April 2024, the AI was cleared to actually fly the jet in flight, with a safety pilot monitoring rather than controlling.

That last phase — in-air autonomy with a human observer — is where VENOM moves from theory to reality.

The Human-on-the-Loop: Safety at the Core

F-16 leading a formation of futuristic collaborative combat aircraft (ccas) in flight
The ai-capable f-16 paves the way for future air combat, leading formations of collaborative combat aircraft.

One of the most important concepts in VENOM — and one that gets misrepresented in popular coverage — is the distinction between “human-in-the-loop” and “human-on-the-loop.”

Human-in-the-loop means a human approves every action before it happens. Human-on-the-loop means the human monitors the AI’s decisions in real time and retains the authority to intervene — but the AI acts independently unless overridden. VENOM uses the latter model during autonomy testing.

The safety pilot seated in the F-16D isn’t passively watching. They’re evaluating the AI’s decision-making against pre-defined safety parameters, ready to immediately reassert control if the system behaves unexpectedly. This isn’t a failsafe tacked on for optics — it’s fundamental to the program’s methodology. DARPA has been explicit: the VENOM aircraft will never fly without a human component.

This human-machine relationship is itself a subject of ongoing research. How much does a pilot trust an AI flying their aircraft? How does that trust develop over time, through data, through demonstrated reliability, through hours logged watching an algorithm make correct calls? These psychological dimensions of human-machine teaming are as important to VENOM’s mission as the raw engineering.

From X-62A VISTA to VENOM: How We Got Here

Conceptual image of the darpa venom autonomy kit hardware with glowing ai circuits
The core of project venom: a sophisticated autonomy kit integrating advanced ai into the f-16’s systems.

VENOM didn’t emerge from a vacuum. Its most direct predecessor is the X-62A VISTA (Variable In-flight Stability Aircraft) — another modified F-16 that DARPA used to prove a fundamental point: an AI can fly a fighter jet.

The X-62A VISTA made headlines in 2023 when it became the first AI-controlled fighter jet to engage in simulated aerial combat against a human-piloted aircraft. It demonstrated that AI could manage the extreme physical and tactical demands of a dogfight — pulling aggressive maneuvers, making targeting decisions, reacting to an opponent faster than any human could.

VENOM builds directly on those lessons, but with a more structured objective. Where VISTA proved that AI could fly and fight, VENOM is designed to prove how AI capabilities can be rapidly developed, tested, validated, and transitioned into operational platforms at scale. Six aircraft instead of one. A repeatable modification process. Faster iteration cycles. The leap from VISTA to VENOM is the leap from proof-of-concept to a scalable program of record.

The Future of Air Combat: CCAs and the Loyal Wingman Concept

Everything VENOM is doing has a destination: Collaborative Combat Aircraft (CCAs), the Air Force’s term for the autonomous uncrewed platforms that will fly alongside crewed fighters as “loyal wingmen.”

Picture this scenario. A human pilot in an F-35 crosses into contested airspace. Flanking them are three AI-controlled CCAs — lightweight, expendable, and capable of acting as sensors, decoys, or weapons platforms. The human sets the mission parameters and retains decision authority over lethal actions. The AI handles navigation, formation keeping, threat response, and data relay at machine speed. If one CCA is targeted, it doesn’t matter the way losing a crewed aircraft would. The human pilot’s survivability improves. The enemy’s radar picture becomes more complex. The mission gets done.

This is the operational vision that VENOM is working toward. The autonomy kit technology, the validated AI models, the human-on-the-loop protocols — all of it feeds directly into CCA development programs already underway, including Skyborg, the Air Force Research Laboratory’s autonomous aircraft initiative.

Why This Changes the Strategic Equation

The implications go beyond tactics. Autonomous wingmen fundamentally alter the cost-exchange problem in air combat. A peer adversary fielding surface-to-air missiles worth $1–3 million per shot can currently destroy a crewed aircraft worth $80–100 million — and kill the pilot. A fleet of capable, AI-controlled CCAs changes those calculations dramatically, especially when combined with the intelligence and situational awareness advantages that AI can generate in real time.

Maintaining air superiority against sophisticated adversaries like China and Russia increasingly requires solutions that scale faster than the pilot training pipeline allows. AI-capable aircraft are one of those solutions.

Challenges That Can’t Be Ignored

VENOM’s progress is genuinely impressive, but intellectual honesty demands acknowledging the obstacles that remain. Three stand out.

Technical reliability in dynamic environments. An AI that performs flawlessly in simulation may encounter edge cases in real combat that training data never anticipated. Ensuring robust performance across every possible scenario — weather, electronic warfare, novel adversary tactics — is an unsolved problem.

Cybersecurity and adversarial manipulation. An AI-controlled aircraft is a software system, and software can be attacked. An adversary who understands the AI’s decision model could theoretically craft inputs designed to confuse or manipulate it. Hardening these systems against adversarial AI attacks is an active area of research with no easy answers.

Ethical and accountability questions. Even with a human-on-the-loop, the question of who is responsible when an autonomous aircraft makes a lethal mistake is deeply complex. International humanitarian law was written assuming human decision-makers. The push toward autonomous weapons systems forces legal and ethical frameworks to evolve in real time. DARPA and the Air Force have committed to maintaining meaningful human oversight — but “meaningful” is a word that will face scrutiny as these systems mature.

These challenges don’t invalidate VENOM’s mission. They define the work that still needs to be done.

Frequently Asked Questions

What does VENOM stand for in the Air Force program?
VENOM stands for Viper Experimentation and Next-gen Operations Model — Autonomy Flying Testbed (VENOM-AFT). “Viper” is a common nickname for the F-16 Fighting Falcon.

How many F-16s are being converted for VENOM?
Six F-16D aircraft are being modified with the VENOM Autonomy Kit for testing at Eglin Air Force Base, Florida.

Is there always a human pilot in the VENOM F-16?
Yes. During all autonomy testing, a safety pilot is seated in the aircraft and can take manual control at any moment. DARPA has confirmed the aircraft will never fly without a human component.

How does VENOM relate to the X-62A VISTA?
The X-62A VISTA was a predecessor program that proved AI could fly and engage in simulated aerial combat in an F-16. VENOM builds on those lessons with a more scalable, repeatable process designed to rapidly develop and transition AI capabilities to future aircraft.

What are Collaborative Combat Aircraft (CCAs)?
CCAs are autonomous uncrewed aircraft designed to fly alongside crewed fighters as “loyal wingmen.” The technology and data generated by VENOM are directly informing CCA development programs like Skyborg.

When did VENOM begin in-air autonomy testing?
In-air autonomy testing — where the AI actually flies the jet with a monitoring safety pilot aboard — began in April 2024, following piloted flights with the autonomy kit engaged that started in late 2023.

The Viper Gets a Digital Brain

DARPA’s VENOM program is one of those rare defense initiatives where the implications match the ambition. Six modified F-16Ds flying at Eglin Air Force Base aren’t just testing software — they’re validating a new model for how air forces will fight. The VENOM Autonomy Kit, the human-on-the-loop safety framework, and the deliberate progression from simulation to flight have already demonstrated that AI can be trusted to fly a combat aircraft. The next step is scaling that trust into a doctrine.

The journey from the X-62A VISTA to Project VENOM to fully operational Collaborative Combat Aircraft is a coherent path, not a collection of disconnected programs. Each step builds on the last, and each flight adds data that moves the needle toward a future where human pilots are force multipliers rather than single points of failure. For anyone tracking where military technology is headed — and if you’re the kind of curious mind who can’t resist a fascinating story at the intersection of science, defense, and the future — VENOM is a program worth watching closely.

Categorized in:

Combat Aviator,

Last Update: July 17, 2026