More than a third of the computer systems that guide 45,000 daily flights across the United States have been rated “unsustainable” by federal auditors, a finding that has pushed the Federal Aviation Administration into its most ambitious technology overhaul in decades. Now, as the agency opens the door to outside vendors, one of the most closely watched questions in government tech is whether Palantir Technologies and its AI platform will play a central role in what comes next.
Decades of delay, quantified
The urgency is not theoretical. A Government Accountability Office report published in April 2024 found that 51 of the FAA’s 138 critical systems were classified as “unsustainable” in the agency’s own 2023 risk review. Another 54 were labeled “potentially unsustainable.” For some of the worst-rated systems, projected replacement timelines stretched 10 to 13 years, a gap the GAO called a serious operational and safety risk.
Those numbers land against a backdrop that air travelers and controllers already feel. Staffing shortages at key facilities have forced the FAA to slow traffic into major airports, and a string of alarming runway near-misses in 2023 and 2024 drew congressional scrutiny. While staffing and technology are separate problems, they share a root cause: an infrastructure built for a different era straining under modern demand.
The FAA’s last large-scale modernization attempt, the Next Generation Air Transportation System (NextGen), launched in 2007 with a target completion around 2025. A 2023 Department of Transportation Inspector General audit found that many NextGen programs ran years behind schedule and billions over budget. That history looms over every new promise the agency makes.
A new program, a hard deadline
Transportation Secretary Sean P. Duffy and FAA Administrator Bryan Bedford have staked their credibility on a successor effort branded America’s Brand New Air Traffic Control System, or BNATCS. The FAA’s program page describes BNATCS as a multi-year push to replace core traffic management tools, consolidate functions, and improve resilience across the national airspace, with a target delivery date at the end of 2028.
To build political momentum, Duffy hosted what the Department of Transportation called the first Modern Skies Summit in spring 2025, a forum where industry vendors and federal officials previewed new software and discussed how to retire legacy platforms that, in some cases, still rely on decades-old code. Bedford used the event to frame the initiative as a generational overhaul rather than an incremental patch.
The first concrete procurement step is already in motion. The FAA launched a vendor challenge to replace the Traffic Flow Management System (TFMS), the tool that coordinates flight schedules and ground delays nationwide. The replacement effort, called Flow Management Data and Services (FMDS), invites private-sector competitors to propose modern alternatives. Bedford tied the challenge directly to the age of existing technology and its drag on system performance, positioning FMDS as an early proving ground for the speed the agency hopes to sustain across the broader BNATCS program.
Where Palantir fits
Palantir Technologies, the data analytics firm known for defense and intelligence work, has an existing commercial relationship with the FAA. Federal Procurement Data System records show contract awards listing Palantir as a vendor and the FAA as the contracting agency. Those records confirm a prior business tie but do not specify the scope, dollar value, or technical focus of the work in enough detail to confirm any AI-specific involvement in air traffic operations.
That limited foothold matters because BNATCS is expected to attract a wide field of bidders. Large defense contractors with long FAA histories, such as Raytheon (now RTX) and Leidos, have deep experience integrating complex aviation systems. Palantir’s competitive edge would more likely lie in narrowly scoped analytics, decision-support, or data-integration contracts, areas where its Foundry and AIP platforms have gained traction across other federal agencies.
As of May 2025, no FAA statement, contract announcement, or vendor challenge document names Palantir as a selected or preferred bidder for BNATCS or the FMDS replacement. Palantir executives have not made public statements about the FMDS challenge or the broader program. The company’s presence in FAA procurement records establishes that it is known to the agency and capable of operating under federal contracting rules, but it does not confirm a pending deal.
The AI question no one has answered
Beyond the vendor horse race, a more fundamental question remains unresolved: how, exactly, the FAA plans to use artificial intelligence in safety-critical airspace operations. Neither the DOT’s summit materials nor the BNATCS program page describe specific AI capabilities, model architectures, or testing and certification protocols.
The possibilities range widely. AI could serve as a back-end analytics layer, crunching weather and traffic data to recommend more efficient routes. It could function as a real-time decision-support tool, flagging conflicts for human controllers to resolve. Or, in a more aggressive scenario, it could automate certain low-risk routing decisions entirely. Each option carries different safety, labor, and regulatory implications, and the National Air Traffic Controllers Association (NATCA), the union representing roughly 14,000 controllers, has historically insisted that technology augment human judgment rather than replace it.
Until the FAA publishes technical requirements or certification guidance for AI in operational air traffic management, any vendor’s claims about what its platform can do in that environment remain aspirational.
Why the 2028 deadline faces long odds
The end-of-2028 target is politically appealing but operationally daunting. The GAO audit noted that some system replacements already carry timelines of a decade or more. Legacy software is deeply embedded in controller workflows, and the FAA’s certification process for safety-critical systems is deliberately slow, designed to prevent the kind of failures that could cost lives.
The agency must also keep the current system running while building its replacement, a constraint engineers sometimes compare to rebuilding an airplane in flight. Procurement bottlenecks, integration testing, and the sheer number of facilities that need upgrades all work against compressed schedules. Whether the vendor-challenge model and commercial partnerships can overcome those structural headwinds is the central bet of the BNATCS program.
History suggests caution. NextGen’s experience showed that political will at the top does not automatically translate into on-time delivery when the work involves thousands of interconnected systems, strict safety mandates, and a workforce that must be retrained on every new tool.
What to watch as contracts take shape
The FMDS vendor challenge is the nearest milestone. Its outcome will signal how aggressively the FAA is willing to move away from incumbent contractors and toward newer entrants, including AI-focused firms like Palantir. Equally telling will be whether the agency structures BNATCS as a handful of large, integrated contracts or breaks it into modular procurements that let specialized companies compete on narrower tasks.
For now, the record supports a clear narrative: the FAA is under documented, auditor-confirmed pressure to modernize infrastructure that is failing in real time. It has launched a branded overhaul with a politically charged deadline and is experimenting with new procurement tools to accelerate the work. AI-focused vendors, Palantir among them, are logically positioned to compete for pieces of that effort. But until the agency publishes specific awards or technical documentation, the question in the headline remains exactly that: a question the FAA has not yet answered.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
