This article responds to Skift’s “AI Isn’t Ignoring Airlines. It Just Can’t Read Their Fares.” (June 17, 2026) and PhocusWire’s “Skyscanner launches AI travel discovery, road trip planning tools” (June 30, 2026).
The signal: discovery is becoming easier to machine-read
Skift reports Adobe’s analysis of more than 8 million U.S. travel-site visits: AI-source traffic rose 194% year over year in May, and those visits were longer, less likely to bounce, and more engaged. The traveller takeaway is not that every AI answer is reliable. It is that a family may arrive at a travel page already carrying a machine-generated shortlist.
PhocusWire describes Skyscanner’s natural-language Explore with AI, comparison of price, duration, weather and destination feel, and a road-trip planner that suggests routes, rentals and stops. Early testing reportedly saw 60% of users click through to generated flight options. Discovery can now move quickly from a vague wish to a plausible route.
The family problem: plausible is not executable
“Kyoto and Osaka in six days” sounds simple until the plan has to respect the arrival gateway, hotel base, luggage, train reservations, opening days, weather, a child’s energy and the departure airport. A beautiful answer can still hide a cross-city day, a tight transfer, or an assumption that was never confirmed.
Try the handoff test before you confirm anything
Take the AI answer and ask four practical questions:
- Can we see the route? Put each anchor on a day and group places that actually belong together.
- Can we see the movement? Name the station, airport, transfer, walking stretch or local transport leg instead of writing “nearby”.
- Can we see the trade-off? Mark what gets removed when the family needs a shorter day, a rainy-day alternative or a later arrival.
- Can we see what remains unconfirmed? Separate a proposed stop from live opening hours, seat availability, price, cancellation terms and provider confirmation.
This is deliberately not a ranking exercise. It is a traveller-control test: can the family edit the plan without losing the reasoning that made the route workable?
Where Alfred fits
Alfred turns traveller details, destinations, dates and pace into a structured, editable itinerary. It gives the family a place to inspect the proposed route, adjust the balance between anchors and breathing room, and keep alternatives visible. That is useful precisely because it keeps planning separate from provider confirmation.
Alfred does not promise that a public page can confirm a live train, attraction slot, fare or hotel term. Those details still belong to the relevant provider. The traveller remains in control of the final choice and booking handoff.
A practical next step
Start with the family’s actual arrival and departure points, ages, pace and must-dos. Then ask Alfred for an editable route and apply the handoff test before opening provider pages. If the route cannot explain its movement and assumptions, it is not ready to book yet.