Google just announced three new travel features for AI Mode in Search: flight price tracking, points and miles redemption viewing, and end-to-end hotel booking. The capabilities are genuinely useful—AI Mode can now track flight prices across more than 300 partner airlines and sites in over 180 countries, show you reward costs for flights and hotels, and let you book directly through integrated partners.
The execution is polished. The user experience is exactly what you'd want from an agentic interface. But as someone who watches platform power dynamics closely, this rollout is less about AI innovation and more about Google tightening its grip on the travel booking funnel.
The features: actually good UX
Let's be fair first. These are real improvements.
Flight price tracking brings Google Flights' popular alert system into the conversational flow. You describe your trip, AI Mode shows options, and you can say "track these flight prices for me." No context switch, no separate interface. You get email alerts when prices drop. It's the kind of continuity that makes agentic AI feel less like a novelty and more like a genuine workflow upgrade.
Points and miles display is the feature road warriors have wanted for years. Ask AI Mode to show flights from Atlanta to Miami using AA miles for specific dates, and it surfaces redemption costs directly from partner data. Same for hotels—tell it you want to use Hilton points, and it shows you the cost. Launch partners include Alaska, American, Choice, Hilton, and Wyndham, with Accor, Flying Blue, Hyatt, LATAM, and Lufthansa coming soon.
Hotel booking closes the loop. Find a hotel in AI Mode, ask to complete the booking, select "Continue on Google" from the booking options, pick your room type, review cancellation policies, and pay via Google Pay. The merchant of record is still the hotel or OTA, so customer service stays with them. Partners include Booking.com, Expedia, Hilton, IHG, Marriott, and others.
From a pure product perspective? This is competent execution. The conversational continuity is legitimately better than bouncing between tabs.
The architecture: classic Google platform play
But zoom out and the pattern becomes clear. Google isn't just adding features—it's inserting itself deeper into the transaction path.
The partner gatekeeping
Notice the phrasing: AI Mode shows costs "based on the latest data directly from our partners." To get your airline or hotel chain's redemption rates surfaced in AI Mode, you need to be a Google partner. The launch list is telling: major chains and OTAs, the players who can afford integration costs and have negotiating leverage.
This isn't technically lock-in, but it's platform control. If you're a boutique hotel chain or a smaller loyalty program, you're not on the initial partner list. Your redemption rates won't show up when users ask AI Mode. Over time, that's a discoverability tax.
The booking flow capture
The hotel booking feature is where the platform dynamics get most visible. Users complete bookings "securely via Google Pay." The hotel or OTA is the merchant of record, but Google owns the payment rails and the final conversion step.
Google gets:
- Payment processing fees (even if small)
- Complete transaction data (who books what, when, at what price point)
- User behavioral data (which prompts lead to bookings, which hotels convert)
- Lock-in to Google Pay as the default payment method in AI Mode
The hotel gets the booking. Google gets the platform leverage.
The uncomfortable questions
Here's what Google's post doesn't discuss:
How does ranking work? When AI Mode shows you flight or hotel options, what determines the order? Is it purely price and relevance, or do partner relationships influence prominence? The post says AI Mode will "show you the best options available," but "best" is doing a lot of work there. We've seen this movie before with Google Shopping and hotel search—algorithmic opacity plus partner relationships creates inherent conflicts of interest.
What's the rev share? When you book a hotel through AI Mode via a partner like Booking.com or Expedia, who gets what cut? The post is silent on economics. Given that OTAs typically take 15-30% commissions from hotels, and Google is now inserting itself into that flow, the margin dynamics matter. Is Google taking a slice of the OTA's commission? Negotiating direct deals with hotel chains? Users deserve transparency here.
Where's the neutral ground? The flight tracking feature pulls from "more than 300 partner airlines and travel sites," which sounds comprehensive—until you realize Google controls which partners get included and on what terms. There's no way for a user to verify completeness or neutrality. You're trusting Google's partner network is genuinely comprehensive, not strategically curated.
Why this matters for AI ecosystems
This rollout is a template for how platform companies will monetize agentic AI. The pattern:
- Build a genuinely useful conversational interface
- Integrate "partners" who provide data and fulfill transactions
- Own the query layer, the ranking logic, and the payment rails
- Extract value through payment fees, data access, and partner negotiations
- Maintain that you're "just" providing user value
It's not wrong, exactly. Users do get value. But it's also not neutral infrastructure. It's platform power dressed up as helpfulness.
Compare this to how ChatGPT or Claude handle similar queries. They'll give you advice, help you think through options, even draft search queries—but they don't own the booking flow. They can't. They're not integrated with payment rails and partner networks. That limitation is also a kind of neutrality.
The builder's dilemma
If you're building AI-native travel tools, Google's announcement is a competitiveness preview. To match this UX, you need:
- Real-time pricing APIs from hundreds of airlines and booking sites
- Loyalty program integrations with every major chain
- Payment processing infrastructure
- Trust and brand recognition to make users comfortable booking through your interface
That's not a startup go-to-market. That's incumbent infrastructure leverage.
The open question is whether there's room for specialized agents that don't try to own the full stack. An AI that helps you think through tradeoffs, surfaces options from across the web, and then hands you off to book wherever you want—no payment capture, no partner gatekeeping, just genuine assistance.
That's a harder business model. But it might be a more honest one.
What to watch
Google is rolling out hotel booking in the U.S. first, with the feature becoming available "over the coming weeks." Pay attention to:
- Ranking transparency: Do partner hotels or OTAs get preferential placement? Does paying Google for ads influence AI Mode results?
- Rate parity: Are the prices shown in AI Mode actually competitive with booking directly or through other channels?
- Data usage disclosures: What does Google do with booking behavioral data? How does it feed back into ad targeting or other products?
- Partner exclusivity: Do partnership terms prevent hotels or airlines from integrating with competing AI platforms?
The technology is impressive. The UX is legitimately good. But the platform dynamics are worth scrutinizing. Agentic AI isn't just a new interface—it's a new control point. And Google is moving fast to own it.