This May, high above the Baltic Sea, a Swedish Gripen E fighter jet sliced through the air, with no pilot at the controls.
During two exercises conducted in May and June, the jet flew under the command of AI software developed by the high-flying defense startup Helsing. A safety pilot sat in the cockpit, but didn’t need to intervene. The test flights marked a milestone for the Munich-based defence tech company and a potential turning point for European military aviation.
Stephanie Lingemann, senior director at Helsing, called the shift “revolutionary.” Speaking at the company’s Munich office, she said that the company’s Centaur AI system gained the equivalent of one million flight hours in just 72 hours. That’s more than a hundred times what a human pilot might accumulate across an entire career. “You can get to superhuman performance very quickly,” she explained to the Financial Times. “And you are not having to send your pilots into dangerous situations.”
Is The Autonomous Jet Fighter Era Upon Us?
Helsing’s autonomous Gripen flights represent the leading edge of a growing global race to develop uncrewed fighter aircraft. Militaries around the world — including the United States, China, and Russia — are pursuing AI-powered jets and “loyal wingmen,” drones that support piloted aircraft in combat.
Yet while the vision of autonomous air combat has waxed and waned since the Cold War, something feels different this time. The war in Ukraine has accelerated Europe’s defence industry and intensified investment in AI systems that can adapt quickly and act decisively.
“This is a paradigm shift in air combat worldwide,” wrote US Air Force Colonel Kevin Anderson in a recent article for NATO’s Joint Air Power Competence Centre.
Fighter jets are expensive to build, maintain, and fly. AI could lessen the burden by removing humans from the equation. In the United States, the Air Force is modifying F-16s through its “Project Venom” to train AI systems for eventual deployment in drones. These systems are envisioned as part of mixed human-machine teams, with uncrewed fighters supplementing or replacing crewed jets.
In Europe, Helsing’s ambitions align with a broader push for defence autonomy. The company, now valued at €12 billion, has grown rapidly since its 2021 founding, supported by investors like Spotify’s founder Daniel Ek. Originally focused on AI software for weapons systems, Helsing now also produces hardware like drones and unmanned underwater vehicles.
Humans Still in the Loop — For Now
Despite the excitement, the road to widespread adoption is not without obstacles. Machine learning — the AI technology powering these systems — is still learning how to interpret and respond to the chaos of war.
Military experts stress that the complexity of aerial combat scenarios makes full autonomy risky. “Robotified warfare…is centuries away,” one air force commander said. Even minor changes, like the presence of tires on a runway, can confuse computer vision systems. In other words, AI may dominate in clean simulations but stumble in the dirty, ambiguous reality of warfare.
Then there’s the ethics. Antoine Bordes, Helsing’s vice-president for AI, insists that human decision-making remains central to how the company’s weapons and software are used. “If we don’t do it in Europe, with our own values, it will be done elsewhere,” he said.
But not everyone is reassured. The possibility of fully autonomous lethal drones — the kind of systems that select and strike targets without human input — remains deeply controversial.
Helsing’s armed drones are already active in Ukraine, where the company has agreements to supply 10,000 units. Yet even there, frontline soldiers have criticized the performance and cost of the company’s HF-1 kamikaze drone and its associated Altra software.
Simon Brünjes, who leads Helsing’s armed drone division, acknowledged the limitations of the HF-1 but expressed confidence in its successor, the HX-2, currently being tested in the field. Importantly, Brünjes also drew a line, at least for now. In high-risk environments like Ukraine, where civilian structures often lie near military targets, he believes human judgment is essential. “In such an environment, we want a human to be making the decision,” he said. Neither the HF-1 nor the HX-2 is autonomous, at least not the models supplied to Ukraine’s Armed Forces.
Still, he admitted, “In other scenarios — full-scale war with Russia or China — it’s a different question.”
What Comes Next?
Even as Europe races to modernize its air forces, most experts expect a lengthy transition period. Helsing’s Lingemann said that AI will first augment human pilots, offering assistance with complex maneuvers and threat detection. Over time, she said, roles will shift. “I think we will have decades where we see both. And then gradually operators — as with drones — will switch to different roles.”
But Europe’s strategic pivot is clear. While uncrewed fighter jets may not dominate air forces tomorrow, their arrival appears inevitable. The technology is rapidly improving. The war in Ukraine is serving as both a proving ground and a political catalyst. And financial support is pouring in from investors eager to back the next frontier of defence.
