Transportation Secretary Sean Duffy is pushing to bring artificial intelligence into America’s air traffic control system, unveiling a sweeping modernization plan in spring 2026 that targets more than 4,600 sites for technology upgrades and pairs new automation with revamped hiring strategies through 2028. The effort represents the most ambitious overhaul of the nation’s aviation infrastructure in decades, and it arrives as chronic disruptions at Newark Liberty International Airport have forced a national reckoning with equipment and staffing failures that officials say are putting lives at risk.
What the plan actually calls for
The Department of Transportation’s blueprint, released by the agency, spans four major categories: communications, surveillance, automation, and facilities. Legacy telecom connections would be replaced with fiber, wireless, and satellite links across thousands of sites. Aging radios, voice switches, and radar systems are slated for replacement or expansion. These are not aspirational bullet points. They represent a defined architecture for how information will move between aircraft, control towers, and regional centers.
The Federal Aviation Administration has tied its workforce strategy directly to these technology goals. In an updated hiring plan, the agency cited “technology advancements such as artificial intelligence” as one factor shaping controller and safety inspector recruitment through 2028. Annual hiring targets and total planned positions were published alongside the modernization push. It marks the clearest official signal yet that AI is being factored into federal air traffic planning, not as a distant aspiration but as a near-term variable that could influence how many controllers are trained, where they are posted, and what skills they need.
Two FAA programs already in motion show what modernization looks like on the ground. The Surface Awareness Initiative, part of the FAA’s surface safety portfolio, aims to sharpen situational awareness for controllers and reduce runway incursions, one of the most dangerous categories of aviation incidents. The system combines sensors and software to flag potential conflicts on taxiways and runways before they escalate. Runway incursion numbers have trended upward in recent years, adding urgency to the effort.
Separately, the Terminal Flight Data Manager program is phasing in electronic flight strips to replace the paper ones still used at many control towers. The TFDM rollout, which began at select facilities in 2023, follows an approved timeline with specific milestones laid out in FAA program documentation. That paper strips persist at dozens of towers in 2026 speaks to how far behind the current system has fallen. Moving to digital flight data is a prerequisite for meaningful AI assistance, because algorithms can only analyze and predict based on information captured in machine-readable form.
Newark as exhibit A
Duffy has framed the overhaul as urgent, pointing repeatedly to Newark Liberty International Airport as proof that the status quo is failing. A combination of staffing shortages and outdated equipment has produced a string of operational breakdowns at the airport, including ground stops, near-miss events, and cascading delays that ripple across the national system. The Associated Press reports that Duffy is asking Congress to approve a funding package described by officials as costing “tens of billions,” using Newark’s troubles to build the political case that delay is no longer acceptable.
The strategy carries echoes of past modernization campaigns. The FAA’s NextGen program, launched more than a decade ago, promised a similar technological leap but was plagued by cost overruns, missed deadlines, and scaled-back ambitions. That history looms over the current proposal and gives skeptics reason to ask whether this round of promises will deliver different results.
What remains uncertain
For all its ambition, the plan is thin on several critical details. No official cost breakdown has been released showing how much of the “tens of billions” would go specifically toward AI tools versus traditional infrastructure like fiber lines and radar hardware. Without that granularity, it is hard to judge whether AI is a central pillar of the overhaul or a secondary feature being highlighted for political appeal. No formal budget request or Congressional Budget Office score has been published.
The FAA has not named specific AI systems it intends to deploy, described testing protocols, or offered projected error-reduction metrics. There is no published institutional research from the agency measuring AI’s direct impact on controller workloads, decision accuracy, or incident prevention. That gap matters. Controllers who rely on AI recommendations without understanding their limits could face new categories of error, particularly during high-stress scenarios where split-second judgment is required and accountability for decisions must remain unambiguous.
The hiring projections raise their own questions. If AI tools are expected to let fewer controllers handle higher traffic volumes, the staffing targets should eventually reflect that efficiency gain. If they do not, either the AI integration is further off than the rhetoric suggests or the tools are meant to supplement rather than reduce human staffing. The FAA has not clarified which scenario it is planning for.
Then there is the transition itself. Upgrading more than 4,600 sites will take years, and during that stretch controllers will work across a patchwork of old and new technologies. The plan does not yet detail how data will be shared between upgraded facilities and those still running legacy equipment, or how training will prepare controllers to move between environments without losing proficiency.
Notably absent from the public discussion so far is a detailed response from the National Air Traffic Controllers Association, the union representing the roughly 14,000 controllers who would work alongside any new AI tools. NATCA has long advocated for better staffing and has expressed cautious support for technology that reduces workload, but the union has also insisted that automation must never replace human judgment in safety-critical decisions. How controllers and their representatives shape the implementation will matter as much as the technology itself.
What independent review shows, and what it doesn’t
The strongest evidence backing this story comes from primary government sources. The DOT’s press release establishes the scope and infrastructure targets. The FAA’s hiring plan confirms AI is influencing workforce strategy. And the agency’s program pages for the Surface Awareness Initiative and TFDM provide concrete examples of upgrades already underway with approved timelines and defined safety goals.
What is notably missing is independent technical scrutiny. No third-party aviation safety organization, academic institution, or government watchdog has published an assessment of how AI tools would perform in live air traffic control environments based on sources available as of May 2026. The Government Accountability Office, which has a long track record of scrutinizing FAA modernization efforts and flagged repeated problems with NextGen, has not weighed in on the AI-specific components of this plan. Without that external review, much of the discussion about AI’s benefits and risks rests on agency aspirations rather than empirical evidence.
Where this leaves travelers and lawmakers
The commitments around fiber upgrades, radar replacements, and digital flight data are concrete and backed by official timelines. The AI elements are real in the sense that they are shaping hiring plans and appearing in strategy documents, but they lack the specificity that would let outside observers judge their safety impact or cost-effectiveness. Congressional appetite for a multi-billion-dollar appropriation is untested. No legislation has been introduced, and bipartisan support for spending at this scale is not guaranteed given competing budget priorities. Any significant delay in funding could force the FAA to stretch limited modernization dollars over more years than planners anticipate.
For now, the United States is embarking on a long-delayed infrastructure overhaul with AI positioned as a promising but still loosely defined piece of the puzzle. Until the FAA publishes detailed technical plans, testing results, and independent evaluations, the gap between what has been announced and what will actually change inside a control tower remains wide. Closing that gap will determine whether Duffy’s vision becomes a turning point for American aviation or another chapter in a decades-long story of modernization deferred.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
