The airline already has a code for the delay. The passenger still has a screenshot and a hotel receipt that may not match it. That gap is the product.
Listen to this essay. Audio version (MP3)
The idea mills keep minting a flight chatbot, a claim-letter filler, and a delaycheck.ai. Three names. One object. A line on an itinerary that does not sit next to the cause code the carrier will report, or next to the commitment the carrier already posted.
Greg Isenberg’s stream and the daily micro-SaaS accounts keep splitting the same reader. One post wants a refund radar for merchants. Another wants a photo upload that becomes a manufactured part. A third wants a waitlist text when a slot cancels. Useful pieces. None of them join the boarding pass the traveler already holds to the public file the Department of Transportation already keeps.
The code is not the promise
Most travelers treat a delay as weather. The app says late. The gate agent says maintenance. The email says we are sorry. A hotel charge lands on the card. A week later the cause is a letter in a Bureau of Transportation Statistics file the passenger will never open.
That letter decides two different rights, and the mills keep collapsing them into one slogan.
The first right is a refund. Under 14 CFR Part 260, a covered airline owes an automatic refund if it cancels the flight, or delays it three or more hours on a domestic itinerary or six or more hours on an international itinerary, and the passenger does not accept a rebooking or a voucher. The reason does not matter. The Department has said the refund rule is separate from how a carrier codes the cause.
The second right is care. A hotel, a meal, a rebooking on another carrier. Federal law does not generally require those. What the passenger can point at is the airline’s own customer-service plan, summarized on the Department’s Airline Customer Service Dashboard. Those commitments attach to disruptions the carrier codes as within its control.
Change the code and you change the hotel. You do not change the refund. The product is the join, not a chatbot that says “you may be owed compensation.”
Why this month, not last year
On September 3, 2026, the Department published a final rule amending 14 CFR Part 234. Federal Register citation 91 FR 56588. It takes effect October 19, 2026.
The rule creates a new reporting category, Section 511(b), for ten events named in the FAA Reauthorization Act of 2024. Carriers that report under Part 234 may no longer place those events in the Air Carrier category, the bucket reserved for causes within the airline’s control.
The ten are specific. Aircraft cleaning after a passenger’s death. Aircraft damage from extreme weather, debris, or sabotage. A baggage-system outage the carrier does not control. A cybersecurity attack. An unexpected government-system shutdown. Overheated brakes after an emergency stop. Unscheduled maintenance, including work driven by an airworthiness directive. An onboard medical emergency. Removal of an unruly passenger. An airport closure from volcanic ash, wind, or wind shear.
Some of those were already outside the old Air Carrier bucket. Maintenance, a medical emergency, removal of an unruly passenger, and overheated brakes were not. After October 19 they have a place to go that is not “we caused this.” Dashboard commitments that say “controllable delay” no longer reach them by default.
A crew timeout still sits in Air Carrier. Weather at the airport still does not. The interesting cases are the ones a gate screen can describe two ways. “Mechanical” can be a deferred item the carrier owns, or unscheduled maintenance the new category now holds. The passenger cannot see the code. The carrier can.
That is the shift. Not a vibe about travel being broken. A reporting rule with an effective date, a Federal Register page, and a dashboard the Department already told large carriers to honor.
Two primitives, one wedge
Primitive one is the photo and the paste. The mills have been shipping receipt readers for months. The input here is the thing already on the phone. Boarding pass. Itinerary email. The delay text. The hotel folio if one exists. Flight number. Date. Origin. Destination. Scheduled time. What the airline said.
Primitive two is the public file. The dashboard already lists what each reporting carrier promised for a controllable cancellation and a controllable delay. Part 260 already states the refund clock in hours. After October 19, Part 234 states which ten causes are not Air Carrier. BTS already publishes the cause file, late, by carrier and airport.
The wedge is not a platform. It is a check a stranger will run this week. Paste the itinerary. Photograph the boarding pass. Get back three lines.
- Refund clock. Did the delay cross three hours domestic or six hours international, and did the passenger refuse the rebooking. If yes, Part 260 does not care about the code.
- Care clock. What that carrier posted on the dashboard, and whether the words on the delay text look like Air Carrier or like one of the ten.
- Mismatch. The text said crew. The later code said maintenance. Or the text said mechanical and the code, when it posts, said Section 511(b).
Charge after the mismatch, or do not charge. A free first read is the whole top of the funnel. If you cannot get a stranger to paste one itinerary this week, you do not have a company. You have a policy thread.
The map is the company
The first hundred checks are a service. The ten thousandth is a dataset.
What compounds is not another dashboard of flight status. Flight status already exists. What compounds is the labeled join: stated reason, dashboard commitment, Part 260 threshold, and the cause code that later lands in the BTS file. Carrier. Airport. Hour. Whether the hotel was offered. Whether the refund arrived without a form.
That map has a buyer who is not the passenger. A corporate travel desk that wants to know which carriers’ “mechanical” texts become Section 511(b) codes. A card issuer that wants to know which hotel charges were care the carrier had already promised. A plaintiff’s shop that wants the pattern, not the anecdote. You do not sell them a login to a status page. You sell them the mismatch rate.
Month one is one carrier and one metro. Paste only. No accounts. Month two adds the hotel folio as a second document on the same record. Month three is the ugly internal scoreboard: flights, stated cause, coded cause, refund cleared or not. That scoreboard is the seed of the business sale. The consumer check stays free or cheap. The map is the price.
Models draft. People file.
A refund request to the airline can be a draft. A complaint to the Department’s aviation consumer site is a filing. So is a chargeback that says the carrier broke a posted commitment. So is any letter that asserts a legal right in the passenger’s name.
The model reads the itinerary and the dashboard. It marks the hours. It marks whether the words on the delay text sit in Air Carrier or in the ten. It drafts the paragraph. A person reads it and sends it. If the product sends the complaint on its own, it is not a checker. It is an unlicensed claims shop with a timer.
The same rule covers the edge cases. Unscheduled maintenance is not a slogan a model should swear to. The passenger does not have the maintenance log. The honest output is a label: the text is consistent with a 511(b) event, or it is not, and here is the commitment that applies if the carrier later codes it as Air Carrier. Uncertainty is a feature. False certainty is how the product gets the airline’s lawyer a free exhibit.
Why this cut, not the last one
The leakage essay was tariffs, seats, and subscriptions. The rebate essay was a nameplate. The recall essay was a label. The short-pay essay was a contractor packet. The bill essay was an EOB against a hospital file. The tax essay was a notice against the roll. The utility essay was a statement against a filed tariff. The housing essay was a unit against a score.
This one is the itinerary version of the same pair, with a clock. A document the traveler already has, plus a public commitment file the Department already forces onto a page, joined in the seventeen days before the new cause code goes live. After October 19 the same join still works. It just has a sharper wrong answer available to the carrier.
The noticing used to require a forum thread and a weekend. It now requires a model that can read a boarding pass and a person who will sign the complaint. Recovery still works because the first check costs the traveler almost nothing.
Someone will own the labeled map of stated cause versus coded cause. The mills will keep proposing a new .ai name for each airline PDF. Ignore the names. Join the line. Keep the map.
Will Tygart — Tygart Media.
This is the idea-mill series.
Leave a Reply