FTFuture Technology
AI

The FAA Is Spending 875 Million Dollars on AI to Fix US Air Traffic Control

· 3 min read · By Nath Connell

Key takeaways

  • The FAA is investing 875 million dollars in AI-based decision support tools for air traffic controllers
  • The system will provide real-time conflict prediction over five to fifteen minute windows and automated sector handoff assistance
  • The FAA currently has approximately 3,000 fewer certified controllers than it considers optimal
  • A near-miss incident at Reagan National Airport in January 2026 killed 67 people, increasing pressure for modernisation

America's air traffic control system is, by most accounts, badly overdue for an upgrade. Controllers are working with software and communication systems in some cases decades old, staffing shortages have reached critical levels, and a series of close calls at major airports over the past two years has kept the issue in the news. Now the Federal Aviation Administration has unveiled its answer: 875 million dollars worth of AI.

The plan centres on a new AI-based software programme designed to help air traffic controllers do their jobs more safely and efficiently. The system is intended to process vast amounts of real-time data, including aircraft positions, weather, flight plans, and airspace constraints, and present controllers with clearer situational awareness and better-prioritised information. The FAA is framing this as a tool to assist human controllers rather than replace them, which is the right framing both technically and politically, given the extreme sensitivity around anything that touches aviation safety.

What the System Actually Does

The specifics of the AI rollout are being implemented in phases. The first phase focuses on what the FAA calls decision support: giving controllers better visualisation of potential conflicts and more accurate predictions of where aircraft will be in the next five to fifteen minutes. This kind of short-term trajectory prediction is exactly the kind of computationally intensive task where modern AI models genuinely excel, and where a reduction in cognitive load for controllers can have real safety benefits.

The second phase involves automating some of the routine handoffs and coordination between control sectors, which currently require controllers to make a high volume of radio calls and data entries. Reducing that routine workload could allow controllers to focus more attention on the genuinely complex situations that require human judgement.

The 875 million dollar figure covers development, integration, and deployment across a significant portion of the FAA's en route centres and terminal facilities. That is not a small number, though it looks modest compared to the scale of the problem. The FAA's wider NextGen modernisation programme has been running since the early 2000s and has consumed tens of billions of dollars with mixed results. The concern among some aviation industry observers is that this AI investment becomes another long-running project that delivers incremental improvements rather than the step change the system needs.

The future, in 3 minutes a day. The biggest tech story explained every morning, free. Get the briefing →

The Staffing Problem That AI Cannot Solve

Here is the tension worth sitting with. The FAA currently has around 3,000 fewer fully certified controllers than it considers optimal, and training a new controller takes roughly three years from enrolment to certification. AI can help the controllers who are there work more safely and with less cognitive strain. It cannot create more of them.

The close-call incidents that have made headlines, including the alarming near-miss at Reagan National Airport in January 2026 that killed 67 people, were products of multiple compounding factors, some of which AI tools might have helped with. But systemic understaffing means individual controllers are handling workloads that stretch beyond what the system was designed for. A smarter interface helps. It does not change the underlying arithmetic.

There is also the question of trust and adoption. Air traffic controllers are a highly skilled, highly unionised workforce with legitimate reasons to be cautious about AI systems they did not design and cannot fully interrogate. The FAA will need to work closely with the National Air Traffic Controllers Association to ensure the tools are actually used effectively rather than being worked around.

None of this makes the 875 million dollar investment a bad idea. Decision support systems have a solid track record in aviation, and well-implemented AI tools genuinely can reduce errors and improve throughput. But it is worth being clear-eyed that this is one piece of a much larger and more complex puzzle, not a solution on its own.

Sources

More from Future Technology