.jpg)
AI flies fighter jet in real combat trials, marking breakthrough for Saab and Helsing. Artificial intelligence isn’t just assisting pilots ...
 |
| AI flies fighter jet in real combat trials, marking breakthrough for Saab and Helsing. |
AI vs Human in the sky
In the United States, the Defense Advanced Research Projects Agency (DARPA) ran the groundbreaking Air Combat Evolution (ACE) program. Taking place at Edwards Air Force Base in California, the tests used the X-62A VISTA — a modified F-16 controlled by AI.
Over multiple sorties in 2023 and 2024, the AI pilot engaged a human-piloted F-16 in within-visual-range dogfights. The AI handled aggressive maneuvers, target tracking, and evasion with no manual intervention from the safety pilot onboard. The program was run jointly by DARPA, the Air Force Research Laboratory, and the U.S. Air Force Test Pilot School. Officials confirmed the system “performed safely and effectively,” marking the first known live AI-human dogfight on record.
Across the Atlantic, Sweden’s defense company Saab and German AI firm Helsing launched Project Beyond, the first European live test of an AI-piloted fighter jet. Conducted above the Baltic Sea in June 2025, the trials saw the AI agent “Centaur” take full control of a Gripen E to perform beyond-visual-range engagements against a human-flown Gripen D.
The project was backed by the Swedish Defence Materiel Administration (FMV) and designed to evaluate real-time autonomy, radar fusion, and combat decision-making under BVR conditions. Both Saab and Helsing confirmed that the flights were fully autonomous, with safety pilots only monitoring. It marked Europe’s most advanced step toward AI-augmented combat aviation.
What These Tests Prove
These demonstrations prove that AI can safely and autonomously fly high-performance fighter aircraft, execute complex maneuvers, and make split-second tactical decisions. They do not yet prove that AI consistently outperforms trained human pilots across all engagement types. Many metrics, including kill probability, energy efficiency, and adaptive behavior, remain classified or under evaluation.
Still, the fact that the AI pilots at Edwards AFB and over the Baltic Sea completed live engagements without intervention is a seismic milestone in aerospace development. As Air Force researchers noted, AI has now moved from simulation to reality.
Edwards Air Force Base was chosen for its unmatched test infrastructure and restricted airspace, critical for live autonomy trials. The ACE project is part of DARPA’s broader effort to push AI into mission-critical decision loops under human supervision. Meanwhile, Sweden’s Project Beyond signals Europe’s determination to remain competitive in defense AI — combining the agility of a commercial AI firm like Helsing with Saab’s engineering heritage. Together, these efforts represent a transatlantic consensus: autonomous air combat is no longer science fiction — it’s a new operational frontier.
The strategic implications are profound. AI-piloted aircraft promise faster reactions, reduced pilot fatigue, and missions that no human could survive — from high-G maneuvers to deep-strike operations in denied environments. But they also raise urgent ethical and legal questions. Who authorizes lethal actions?
What happens when algorithms err? And how do we maintain human accountability in the loop? Experts at CSIS and RAND Corporation warn that without global norms, autonomous aerial combat could escalate conflicts faster than humans can respond. The challenge isn’t just to make AI fly — it’s to make AI accountable.
Next-generation programs, including the U.S. Collaborative Combat Aircraft (CCA) initiative and Europe’s Future Combat Air System (FCAS), plan to pair AI-controlled drones and fighters with human pilots. The future of aerial warfare is hybrid — human creativity directing machine precision. The day AI “wins” a dogfight is not the end of the story. It’s the beginning of a new question: How do we lead, govern, and trust machines that can outthink us — at 30,000 feet?
