Mission

flighthelp exists to be the open, neutral, audited public infrastructure for air-travel facts and passenger rights — schemas, data, and a rules engine that everything in the world touching airline policy, baggage rules, and passenger compensation can be built on.

The mission is structural, not transactional. The project does not sell anything to travelers. It does not run ads. It does not file claims. It does not compare airlines for purchase. It does one thing: it maintains and publishes the dataset and the legal logic that the air-travel ecosystem currently has to build privately, badly, and over and over again.

The reference website at flighthelp.net exists as proof-of-work that the underlying infrastructure is real, current, and usable. The website is not the product. The infrastructure is the product.

Why this needs to exist

Three failures, each decades old, justify the project.

The first failure is informational. Travelers cannot reliably answer simple questions about flying — what their bag allowance is, what number reaches a human, what they are owed when something goes wrong. The information exists, but it is hostile: scattered across airline websites that change without notice, regulator pages written in legal prose, and dozens of fragmented commercial sites with stale data and conflicts of interest. No neutral, canonical source exists. The traveler is left with Google, Reddit, and luck.

The second failure is structural. Every commercial actor in the space — claims companies, OTAs, corporate travel platforms, insurance underwriters, AI assistants — currently maintains its own private, partial, often incorrect representation of the same underlying facts. They each pay independently to scrape the same airline pages. They each hand-encode the same EU regulation, frequently with errors. They each compete on top of an information layer that nobody owns and everyone wastes resources rebuilding. The industry is structurally over-paying for the same broken data.

The third failure is regulatory. The legal frameworks for passenger rights — EC 261/2004, UK 261, ANAC 400, Montreal Convention, US DOT regulations, Canada APPR, India DGCA — exist in legal text but not in code. There is no canonical, versioned, auditable computational representation of these laws. Lawyers and claims companies translate them into private codebases. The translations diverge. Travelers can't audit them. Regulators can't reference them. Courts have no shared technical artifact to point to.

flighthelp's mission is to fix all three by building the infrastructure layer that should have existed twenty years ago: open schemas, an open rules engine, an open and verified dataset, and a free public API that any builder — commercial, academic, governmental, or AI — can use as the single source of truth.

What the mission is not

The mission is not to help travelers directly. Travelers are beneficiaries, not customers. They are helped because every product that touches their travel — their booking app, their AI assistant, their insurance claim, their corporate travel tool — gets better when it can build on flighthelp. The reference site at flighthelp.net is a free service to travelers, but it is a side effect of the mission, not its center.

The mission is not to compete with anyone. Not AirHelp, not Skyscanner, not Wikipedia. Commercial claims services, comparison sites, and consumer apps are downstream consumers of the infrastructure, not competitors. The most successful version of this project is one where every existing commercial player ends up using the schemas and engine, and the project is invisible to the traveler because it is everywhere.

The mission is not to make money for its founders. The non-profit governance is structural — not a brand position. The project cannot be acquired. The dataset cannot be locked up. The rules engine cannot be made proprietary. These constraints are written into the governing documents and the licenses so that no future board, no future funding crisis, and no future acquirer can undo them.

Time horizon

flighthelp is built for the long run — measured in decades, not quarters. The model is OpenStreetMap, Wikipedia, Let's Encrypt, the Open Library, Crossref, schema.org. Public infrastructure projects that succeed are not the ones that raise the most funding or get the most press. They are the ones that survive long enough to become load-bearing — that get woven into enough downstream products and institutional workflows that removing them would break too much.

Becoming load-bearing takes ten to fifteen years. The project plans for that horizon. Every decision about governance, licensing, funding, and architecture is made with one question in front of it: will this still make sense, and still be operating, in 2040?

If it would not, it is the wrong decision.