The U.S. Defense Advanced Research Projects Agency (DARPA) and the U.S. Air Force have announced the successful testing of artificial intelligence that controlled a modified F-16 fighter jet.
This marks the first instance of an autonomous control system being used on a retrofitted aircraft of the previous generation.
The test utilized the Viper Experimentation and Next-generation Operations Model (VENOM) kit, which enabled the installation of an AI control system on the F-16 without altering the aircraft’s core software. VENOM is capable of managing flight operations, as well as overseeing sensors and other onboard systems.
During the test flight, a human pilot performed the takeoff, after which control was handed over to the artificial intelligence. The pilot remained in the cockpit, monitoring the system’s performance and ready to intervene if necessary. This approach is referred to as human-on-the-loop: the human observes the AI’s actions and retains the ability to step in, but is not required to approve every decision, unlike the human-in-the-loop model.
According to DARPA, using a production-standard F-16 instead of a purpose-built experimental aircraft simplifies the future scaling of this technology. Multiple upgraded fighters can be used simultaneously to evaluate various AI agents, tactics, and interactions between manned and unmanned aircraft.
The VENOM trials are also linked to DARPA’s Artificial Intelligence Reinforcements (AIR) program. In the long term, such systems are intended to enable pilots to manage groups of unmanned aerial vehicles.
DARPA notes that the application of AI in aviation could assist humans in handling the growing complexity of modern combat missions. However, issues of reliability and trust in autonomous systems under real-world conditions remain critical for further development. The VENOM tests are expected to provide the data needed to transition these technologies from simulations to actual flights.
About the Author
