The most credible AI launches in business aviation this month came from companies that already own the data.

Airframe announcements were thin at this year's NBAA-BACE. Several trade writers noted that AI agents took center stage instead. AvBrief's Julie Boatman singled out two debuts, from WingX and Jeppesen ForeFlight, as examples of AI applied with some discipline.

WingX: from dashboards to agents

WingX builds on the aircraft sales and operations data that Jetnet has collected for decades. Its founder, Christoph Kohler, described the shift as moving from the dashboard era to the agent era. In a live demo, the agents answered questions like where demand is, where money is being spent and who is behind it. One example compared the number of aircraft parked at Las Vegas-area airports during NBAA with prior years, then did the same for the Las Vegas Formula 1 race. FBOs could use that kind of analysis to plan for event surges, and brokers could use it to spot prospects.

The agents draw on more than 15 years of history and stay inside that data sandbox, rather than pulling from the open web.

Jeppesen ForeFlight: Airflow and the "why-chain"

Jeppesen ForeFlight introduced two agents on its new Airflow platform: Trip Feasibility and Ops Briefing. Its SVP of flight operations, Reggie Arsenault, laid out five principles behind them:

  • A context graph, so agents work from aviation data rather than general web content
  • Certified data
  • A "why-chain" that shows how the AI reached a recommendation so humans can check it
  • Safety management system principles built in
  • A human in the loop

ForeFlight is also using AI to condense NOTAMs into a brief view specific to a flight, and says that will show up in its mobile app.

Operations-side use cases

At an NBAA panel on emerging technology, SayFlight founder Toby Benenson described an aircraft-on-ground scenario in which an AI assistant finds a replacement aircraft and drafts a customer notification. Panelists stressed that aviation safety depends on predictable systems, and they drew a line between advanced automation and AI that can learn and change its own behavior.

Floating Fleet AI, an AI scheduling company that also exhibited, has pitched tools for scheduling aircraft and crews and for evaluating incoming trip requests against an operator's network. Its claims of cutting quoting time from hours to seconds come from the company itself.

Why it matters

The most important idea is also the simplest: trust comes from the data and the audit trail. A charter customer will tolerate a chatbot being wrong about a movie recommendation. They will not tolerate it being wrong about a fuel stop or a crew limit. Companies that own certified or proprietary data, and that show their reasoning, are in a stronger position than general-purpose agents.

What to watch

  • Whether these agents move from demos into daily use at charter operators and brokers.
  • How "show your work" features hold up on messy real-world requests.
  • Whether smaller operators can access this kind of tooling at a reasonable price.

Bottom line: The next phase of AI in private aviation looks less like a chat window and more like an assistant that cites its sources.

Sources: