The US Air Force is developing an artificial intelligence-powered chatbot to support student pilot training, marking what officials describe as a logical next step in modernising flight instruction.
According to Air & Space Forces Magazine, the 19th Air Force’s Flying Training Center of Excellence is building an AI tool named “IP GPT”, designed to function as a virtual instructor pilot.
The system will be trained exclusively on aviation manuals, official instructions and flight guidance to provide students and instructors with rapid access to reference procedures and performance insights.
If successful, IP GPT could coach students in simulators, potentially freeing up human instructors for higher-value training tasks and expanding overall capacity.
What sets IP GPT apart from ChatGPT?
Unlike general-purpose large language models such as ChatGPT or Google’s Gemini, IP GPT is being trained on a tightly controlled body of aviation-specific documents. These include Air Force instructions, standard operating procedures, Federal Aviation Administration and International Civil Aviation Organisation guidance, and other flight publications.
Lt. Col. Seth Hoffman, head of the centre’s Detachment 24, told Air & Space Forces Magazine that the objective is precision rather than breadth. “I don’t want it to cue from the whole internet,” Hoffman said, adding that the system should operate within a defined “data pond” of aviation-centric material to avoid confusion or irrelevant information.
Col. Brian Muto, director of the Flying Training Center of Excellence, said adopting such tools is both practical and necessary as younger cohorts enter the service. “If the students are every day on their phone using large language models, they’re going to expect it when they turn 20 and they come to fly the T-7 with us,” he told the publication.
Officials believe the technology is feasible in the near term. Hoffman said advances in AI suggest a working system could be ready within six months. However, he acknowledged that policy and cybersecurity concerns remain significant hurdles. “There are a lot of folks in the cybersecurity realm who are like, ‘Eh, I don’t know about that,’” he said.
In its initial phase, IP GPT would serve as a digital assistant, available before and after sorties to help students review procedures and assess performance. Over time, leaders envision more advanced capabilities, including a data analytics engine capable of reviewing simulator or aircraft data and grading flight manoeuvres against official standards.
Such a system could also assist instructor pilots in preparing training scenarios and conducting debriefs, potentially returning valuable time to instructors while maintaining rigorous oversight.
How AI and data are reshaping India’s aviation boom
India’s aviation sector is expanding rapidly, and airlines are increasingly relying on artificial intelligence and unified data to keep pace without losing control of service quality. The shift is most visible in customer operations, where carriers are automating routine queries, predicting the impact of disruptions, and stitching together fragmented passenger data into a single view.
Air India — now in the middle of a multi-year transformation plan launched in 2022 — has been modernising its service operations, according to Salesforce India’s blog. The airline is using AI features to classify and route cases automatically, recommend next steps for agents, and support proactive service gestures. The company says it handles around 550,000 service cases a month through Salesforce Service Cloud.
Meanwhile, IndiGo, India’s largest airline by market share, is also leaning heavily on automation, according to the same Salesforce blog. The airline says a service team of more than 1,280 representatives handles over 1.2 million cases per month, supported by automated categorisation and a unified service console that pulls data from seven systems.
IndiGo is also using automated feedback capture across the travel journey, reporting around 5 million customer responses annually via Salesforce Feedback Management.
The future of AI in the aviation industry
Looking further ahead, officials suggest AI could support fully qualified pilots as well, providing mission preparation tools through digital flight books and mobile devices.
Human instructors would remain central to training, but AI assistants could augment their role — offering continuous availability and rapid access to approved guidance.
As militaries globally examine the operational potential of AI, the Air Force’s initiative signals a pragmatic approach: applying controlled, domain-specific artificial intelligence to expand training capacity while preserving human judgement and oversight.
Join the conversation
No comments yet
Be the first to share your thoughts!
Sign In to Comment