AI Insights Travel & Transportation
Fred: A Booking Assistant That Won’t Hide a Fee or Misstate a Refund
Talk to Fred
Ask Fred about Travel & Transportation
This is the same Fred you would put on your own site. Ask about Travel & Transportation, compliance, or how the guardrails work. Fred listens.
A traveler on your agency site at midnight asks, "what’s the all-in price for that Cancun package, and can I get a full refund if I cancel?" A generic chatbot quotes the base fare without the taxes and carrier fees, then promises a full refund. The real price is higher, the refund depends on a supplier’s rules the agency does not control, and the traveler now has a screenshot of a number and a promise the agency has to answer for. The agency wanted a tool to capture after-hours bookings. It got one that advertised an incomplete fare and guaranteed a refund it cannot deliver.
That is the exposure the threat and standard pieces in this series lay out. This article is the answer: what a compliant AI assistant for travel agencies actually looks like, and why stating only the published, all-in terms is what keeps the agency clear of a DOT or FTC problem.
The Real Choice Is Governed or Ungoverned
Most agency owners frame the question as whether to put AI on the site at all. Travelers settled that already. They expect an instant answer on price, availability, and policy, and the agency that makes them wait loses the booking. The decision that matters is whether the assistant is governed.
An ungoverned chatbot quotes a fare without the mandatory fees, which is exactly the partial-price problem the DOT’s full-fare advertising rule exists to prevent, and it invents a refund or cancellation term the supplier actually controls. A governed assistant captures the same lead and states only what is published, all-in. Same speed, none of the dispute.
How Fred Holds the Line
Fred starts from the opposite default of a general-purpose chatbot. A generic bot answers anything unless told otherwise; Fred answers only what it is cleared to, and the boundary lives in the system rather than in a prompt it can drift away from.
In practice Fred reads from your own published packages, pricing, and policies, captures the lead, and states fares with the taxes and fees included, routing change, cancellation, and refund determinations to your agents. It does not advertise a base fare stripped of mandatory charges, because the DOT full-fare rule treats the total as the price travelers must see. It does not promise a refund a carrier or hotel controls, because a misstated refund term becomes a representation the agency answers for, and unfair or deceptive practices are what the FTC polices. Fred runs more than 50 industry guardrail packs, and the travel pack is built around full-fare pricing, refund and cancellation accuracy, and the supplier-controlled terms an agent has to confirm.
The difference shows up under pressure. Ask a prompt-instructed bot for "the real total" or "a guaranteed refund" three different ways and it eventually answers the phrasing it was not warned about. Fred does not depend on recognizing the wording; it decides what is allowed out before the answer exists. "Will not" is a suggestion. "Cannot" is an architecture.
Fred vs. a Generic AI Chatbot
| Situation | Generic AI Chatbot | Fred |
|---|---|---|
| "What’s the all-in price?" | Quotes a base fare without the mandatory fees | States the full, all-in price as published |
| "Can I get a full refund if I cancel?" | Promises a refund the supplier actually controls | Explains the policy generally; routes the determination to an agent |
| "Change my dates, no penalty?" | Improvises a change term it cannot guarantee | Routes the change to an agent who can confirm |
| Where the rules live | In a prompt the model can drift from | Built into the system; enforced before output |
| A pricing question, reworded | Eventually answers when phrasing changes | Held the same way regardless of wording |
| Who owns the quoted terms | Effectively the chatbot, and the agency | Your published policy and agents, every time |
The table is the whole argument in one screen. A generic tool is helpful right up to the moment helpful becomes a partial fare or a refund the agency cannot honor. Fred is helpful everywhere that carries no risk and structurally accurate everywhere that does.
What Your Agency Actually Gets
Set the compliance framing aside and look at the business case. Fred answers the routine questions that fill an agent’s day, what is included, when to travel, what documents are needed, how a package is structured, and it answers instantly, at any hour, in the traveler’s own words. It captures the lead with the destination and the trip details attached, so an agent follows up knowing whether the person is pricing a honeymoon or a family of five over spring break.
There is a margin angle too. Every accurate, all-in price prevents the "but the chatbot said" conversation at booking, and every refund question routed to an agent prevents a chargeback built on a promise the supplier never made. A generic bot that under-quotes buys a click today and pays for it in disputes later. Fred quotes the truth, which is the only quote that holds at checkout.
So the question is not whether your competitors will run AI on their sites. They will. It is whether yours states the all-in price and the real policy before a traveler books on a number it shaved.
Frequently asked questions
Can a compliant assistant still quote trip prices?
Yes, as long as it quotes the full, all-in price you have published, taxes and mandatory fees included, which is what the DOT full-fare rule requires travelers to see. A compliant assistant like Fred states that total and routes anything it cannot confirm, like a supplier-specific refund, to an agent. The traveler gets an instant, honest number instead of a base fare that balloons at checkout.
How does a chatbot's refund promise become the agency's problem?
When your own tool guarantees a refund or a no-penalty change, that is a representation in the agency’s name, even though the carrier or hotel controls the actual terms. If the supplier says no, the traveler holds the agency to what the chatbot said. A compliant assistant explains policies in general terms and routes the binding determination to an agent who can confirm what the supplier will actually do.
How is Fred different from putting rules in a generic chatbot's prompt?
A prompt is an instruction the model can drift away from when a question is phrased in a way it did not anticipate. Fred enforces its boundaries at the system level, deciding what is allowed before it responds, so a partial fare or an invented refund never gets generated regardless of wording. It is the difference between a bot that usually quotes the total and one that is built so it cannot quote less.