Here is a genuinely strange place to start a story about an AI pricing threat: the reporter whose work triggered a congressional investigation into Delta’s pricing practices is himself skeptical that the technology will deliver what Delta’s own CEO has promised.
Gary Leff, who writes View From the Wing and has covered airline pricing since 2002, broke the story that put Delta’s AI pricing plans in front of the U.S. Senate. He also happens to think Delta could be overestimating the financial upside, a skepticism he laid out in his own analysis of the CEO’s numbers.
That tension, between a genuine controversy and a genuine doubt about whether the underlying technology is as powerful as advertised, is worth sitting with before accepting either the panic or the hype at face value, especially for anyone whose job is actually paying these fares rather than debating them in a hearing room.
What Delta actually announced
Delta President Glen Hauenstein told investors on the company’s Q2 earnings call that AI-driven pricing, run through a partnership with the Israeli firm Fetcherr, covered roughly 3% of fares at the time of the call, up from 1% earlier in the year, with a plan to reach 20% of domestic pricing by the end of 2025, figures first reported through Delta’s own earnings commentary.
Fetcherr describes its own technology as predicting the next price the way a language model predicts the next word, a transformer-based system plugged directly into fare filing so prices can update near continuously rather than waiting for a human analyst to review a route and adjust it manually. Hauenstein has called early results “amazingly favorable,” according to the same congressional letter that later cited his comments as cause for concern.
CEO Ed Bastian has described the broader opportunity in blunter financial terms: a 2 to 4 point improvement in cost margins over several years, moving Delta from roughly 10% to 15%, which he has characterized as a roughly 50% improvement in profitability worth billions of dollars.
Delta has also expanded AI into functions well beyond pricing, reservations staff use an AI-powered knowledge tool, maintenance planners use AI to forecast needs, crew schedulers use it to anticipate replacement staffing, and the airline’s Baggage AI system improved Atlanta’s mishandled-bag rate by more than 25% year to date.
Delta has said publicly it does not expect an immediate reduction in headcount from any of this, describing the shift instead as reallocating people toward higher-value work rather than eliminating roles outright.
What triggered the fight in Washington
The controversy began in July 2025, when Democratic Senators Ruben Gallego, Mark Warner and Richard Blumenthal sent Bastian a letter warning that individualized, surveillance-based pricing “eliminates a fixed or static price in favor of prices that are tailored to an individual consumer’s willingness to pay,” according to Senator Warner’s own office.
The senators were not reacting to speculation. They were reacting to Hauenstein’s own words, and a follow-up letter from a group of House lawmakers in November laid out the specific language that alarmed them: Hauenstein had told investors that generative AI-powered pricing would ensure airfares are tailored to “that flight, on that time, to you, the individual,” and Fetcherr’s own chief executive had said his company works with “all the data we can get our hands on,” both quotes cited directly in the House lawmakers’ letter to Bastian.
The reaction went beyond letters. Democratic lawmakers Greg Casar and Rashida Tlaib introduced legislation the same month that would bar companies broadly from setting prices or wages using Americans’ personal data, with a provision written specifically to stop an airline from raising a fare after detecting that a traveler had searched for a funeral home or an obituary.
That bill remains pending, and its narrow, almost anecdotal framing says something important about how this fight is actually being understood in Washington: not as a debate over yield management in the abstract, which airlines have practiced for forty years, but as a debate over what happens when that practice gets access to signals about a person’s life that have nothing to do with travel at all.
Senator Gallego put the underlying suspicion plainly when he said Delta appeared to be telling its investors one story and the public a different one, a tension this newsletter’s own reading of the record bears out.
Delta’s answer, on the record
Delta’s response has been consistent and specific. In an August 2025 letter responding to the senators, and again in November responding to the House, the airline stated plainly that “there is no fare product Delta has ever used, is testing or plans to use that targets customers with individualized offers based on personal information or otherwise,” as Delta’s own EVP of external affairs wrote in the company’s public response.
Delta’s position is that its AI pricing tool works on aggregated market data, thousands of flights and fare buckets analyzed at once, not on any individual traveler’s browsing history, device, or personal circumstances, and that the same price is shown to every shopper searching that flight at that moment.
That distinction, aggregated demand data versus an individual profile, is the entire legal and ethical hinge of this dispute, and it deserves to be taken seriously rather than dismissed outright as corporate spin crafted purely for a congressional audience. Nothing in the public record contradicts Delta’s specific claim that no fare has been set using a named traveler’s personal data.
What the record does show is a genuine gap between the caution of that legal language and the ambition of Hauenstein’s own earlier framing to investors, a gap Delta has never fully reconciled in public, and one lawmakers have explicitly asked it to reconcile without yet receiving, in their view, a complete answer.
The mechanism, according to the person who has followed it closest
Leff’s own explanation of what Delta’s system likely does is less dramatic than either the CEO’s promise or the senators’ fear, and more useful for understanding what actually changes. The system reallocates inventory across pricing buckets that already exist, opening more low fares to price-sensitive shoppers on routes and dates where demand data suggests they are searching, while holding higher fares where it suggests they are not, rather than computing a unique number for each individual.
“The airline earns more overall because it discounts where a discount changes behavior and avoids discounting where it does not,” Leff explained in a separate analysis of the mechanism, extending a low fare only where it actually changes someone’s decision to buy.
That is a faster, more granular version of yield management airlines have run since the 1980s, when carriers first began splitting a single flight into a dozen or more fare buckets tied to advance-purchase windows and ticket restrictions, a practice born at American Airlines and refined industry-wide long before anyone called it artificial intelligence, rather than a new category of pricing altogether. It is exactly why Leff, who understands this mechanism better than most people writing about it, remains cautious about Bastian’s 50% profitability figure.
Faster and more granular is a real advantage. It is not automatically the transformational one described on an earnings call.
What this actually means for a corporate travel budget
Here is the part of this story that has gotten the least attention, and it matters more to this newsletter’s readers than the entire personal-data fight in Washington. Airlines have never needed personal data to charge business travelers more, because the booking pattern itself has always been the signal. The Saturday-night-stay requirement that dominated fare structures for decades existed specifically to separate the two groups, business travelers fly out midweek and want to be home by the weekend, leisure travelers are happy to stay over, and airlines priced accordingly with no personal data involved at all.
When low-cost carriers made one-way pricing common enough to break that rule, the industry replaced it with basic economy fare restrictions engineered for the same purpose, as Leff documented in an earlier piece on exactly this shift. Advance-purchase timing does the same work on its own: research on international airline pricing has found that fares rise faster for business travelers than leisure travelers as a flight date approaches, purely because the timing of the purchase itself reveals which group is buying, independent of any personal information about who is doing the buying.
A faster, AI-driven version of this same segmentation does not need to know anything about a specific traveler to extract more from a corporate booking pattern than a leisure one. One-way, midweek, booked inside a two-week window, no Saturday night included, that pattern alone already tells an aggregate pricing model everything it needs to know, entirely within the boundaries Delta insists it will not cross.
A corporate travel program worried about AI pricing should worry less about whether Delta profiles individual travelers and more about how much sharper the airline’s aggregate read of a business trip’s shape is about to become.
The practical question for a travel manager is not whether to trust Delta’s privacy assurances, it is whether the program’s own booking data still looks the way it did two years ago. If a program’s average advance-purchase window has been shrinking, or its share of one-way and midweek tickets has been growing, that program is precisely the shape of demand a faster pricing model will learn to charge more for first, regardless of what data protections apply to any individual traveler inside it.
Programs that have consolidated more of their bookings into predictable channels, with visibility into how far in advance trips are actually being purchased, are in a better position to notice this shift as it happens rather than discovering it eighteen months later as an unexplained rise in average fare. The congressional fight is about whether the line against personal data gets crossed. The budget impact shows up whether or not it does, because the line was never the thing protecting a corporate fare to begin with.
Nobody has said anything about this outside the United States yet
Every airline named in this controversy, and every airline reported to be adopting similar tools, JetBlue with FLYR, Virgin Atlantic and flydubai and Lufthansa with PROS, operates primarily in North America, Europe or the Gulf.
No Latin American carrier has said a word publicly about AI-driven dynamic pricing, which likely means less that the region is behind and more that nobody has asked the question yet in a market where corporate travel budgets are just as exposed to the same booking-pattern logic as anywhere else, midweek trips, short advance windows and one-way tickets being just as common a shape for a business trip in São Paulo or Bogotá as in Atlanta.
A congressional letter is a distinctly American mechanism for surfacing this kind of practice, built on a level of airline financial disclosure and legislative attention that has no direct equivalent in most of the markets this newsletter actually covers. The absence of a letter is not evidence of the absence of the practice, only of the scrutiny, and a regional carrier quietly adopting the same kind of tool without ever facing a congressional question would not be surprising in the least.
About me
I am an entrepreneur with over 20 years of experience at the intersection of tourism and technology. I am co-founder and Chief Business Officer of VOLL, the largest mobile-first corporate travel and expense management platform in Latin America, and a recognized reference in the development of the corporate travel industry.
A Marketing specialist from Fundação Dom Cabral, I serve on the Tourism Council of FecomércioSP and on the Executive Council of the Latin American Association of Corporate Events and Travel Management (Alagev). A frequent traveler and close observer of human behavior in motion, I write and speak about innovation, digital transformation, entrepreneurial leadership, and the future of corporate travel.