Why does travel keep being the first place AI shopping agents get tested in the real world?
How does travel commerce become the first proving ground for agents?
This explores why travel, with its bookings, OTAs and complex itineraries, keeps showing up as the place where AI shopping agents are tested first, and what that testing has revealed so far.
This explores why travel is where AI agents are meeting real commerce earliest, and what has happened there. The corpus doesn't contain a direct argument for why travel was first. It does show that travel is where the predicted effects of agents first became measurable, and that many of the early predictions turned out wrong. The most popular one was that agents would cut out online travel agencies like Expedia and Booking. The 2026 data points the other way. Independent hotels' dependence on OTAs rose to 63.4%, and Expedia's B2B business, which sells its inventory and booking systems to other companies, grew 24% to reach 38% of revenue. OTAs are turning into the settlement layer that agents book through (Are AI agents actually removing online travel agencies from bookings?). OTA stocks rose despite the warnings. One reading is that agents make *discovery* cheap but don't touch the hard parts that come after it: payment, supply contracts, trust and actually completing the booking (Are AI agents actually disintermediating online travel platforms?).
Travel therefore shows something easy to miss. The value doesn't vanish when agents arrive. It moves. The clearest loser so far is the company that sold *attention*, not the company that fulfills the trip. Tripadvisor's hotel revenue fell 23% in one quarter. The explanation offered is that agents rank options against what the traveler asked for instead of looking at paid placements, so the agent keeps the demand signal that ad-funded aggregators used to sell (Are AI agents making attention-based ads irrelevant for aggregators?). The same logic appears outside travel. If agents ignore ads, merchants may move ad budgets into lower prices and better terms (Could agents redirect merchant ad budgets toward customer savings?). Travel also gives agent builders a clear model for making money: start with tiny payment fees and work toward the roughly 12% cut that booking platforms take (How could an AI agent's revenue per transaction grow over time?).
The proving ground has also exposed limits. Travel companies are investing heavily in agentic AI, but most projects are stuck in pilots. The blocker isn't model quality. It's scattered data and ownership across airlines, hotels and distributors (Why is travel's agentic AI investment stuck in pilot mode?). Traveler trust splits in a similar way. Over 90% of travelers are comfortable letting AI brainstorm trips, but comfort drops sharply for high-stakes facts like visa requirements, where a hallucination has real consequences (Do travelers trust AI differently for brainstorming versus visa tasks?). Payments infrastructure also adds strain everywhere. Fraud systems built to read human signals of identity and intent struggle when an agent clicks "buy" (Does agent-initiated commerce break existing payment infrastructure?).
The deeper question travel raises is *whose* agent is doing the booking. Agents built by a platform tend to optimize for the seller, while agents the consumer owns can be set up around the buyer (Does agent ownership determine whose interests get optimized?). Agents funded by advertising carry a built-in pull toward paid placements (Do advertising-funded shopping agents favor paid product placements?). Even a truly consumer-side agent only helps you navigate the existing market. Changing how the market actually works would require seller-side agents negotiating with buyer-side ones (Do one-sided consumer agents actually change market structure?). The takeaway: travel's first lesson is that agents reshuffle *who earns the margin* faster than they change *how people buy*. Ad-driven discovery is collapsing while the OTA plumbing underneath gets stronger.
Sources 11 notes
Goswami's 2026 travel data shows independent hotels' OTA dependence rose to 63.4% even as AI agents emerged, with OTAs retaining merchant-of-record status through API integrations. Expedia's B2B segment grew 24% year-over-year, now 38% of revenue.
Expedia, Booking, and Airbnb posted strong 2026 gains despite disintermediation warnings. Prieto argues AI may cheapen search but leaves payment, supply, trust, and transaction completion intact—functions OTAs control. This implies value may shift rather than evaporate.
As shopping and travel agents handle comparison and ranking based on stated criteria rather than paid placement, they render attention ads irrelevant. Tripadvisor's Hotels segment revenue fell 23% in Q2 2026, with management attributing further decline to structural SEO shifts—a pattern Jarc attributes to agent-mediated ranking.
Jaipuria argues that once agents replace human shoppers, merchants have incentive to convert ad budgets into lower prices and better terms rather than sponsored listings, since agents optimize for value rather than page position. Google's Direct Offers pilot demonstrates this shift in action.
Jaipuria models a three-stage revenue path: starting with sub-0.5% payments commissions, adding 4% affiliate fees, then moving to 12% aggregator take in suitable categories. The blended progression suggests agent economics scale from minimal initial fees toward booking-platform-like margins.
Show all 11 sources
McKinsey and Skift found that travel's fragmented ownership structure and service-focused culture, rather than AI capability, limit agentic use cases to pilot stage. Horizontal copilots scaled easily but delivered diffuse gains.
McKinsey and Skift's survey of 1,002 travelers found over 90% report confidence in AI travel information overall, but comfort drops significantly for high-stakes tasks like visa requirements compared to open-ended brainstorming, driven by concerns about AI hallucinations.
McKinsey argues that payment gateways and fraud engines built around explicit human identity and intent face significant strain when autonomous agents initiate transactions. The firm proposes cryptographically signed mandates as a replacement for human-observable signals fraud systems traditionally relied upon.
MIT research argues that platform-built agents inherit seller incentives by default, while consumer-owned agents can be configured around buyer preferences. The mechanism is principal-agent logic: whoever builds and maintains the agent sets its objective.
Bob Hedges, former Visa Chief Data Officer at MIT, warns that advertising business models create structural incentives for shopping agents to favor paid placements over neutral recommendations, independent of technical design quality.
Choudary argues that consumer-only agents like Muse and Instinct change where buyers enter markets but cannot alter market mechanisms themselves. True structural change requires seller-side agents too, enabling negotiation between represented parties rather than buyer optimization within fixed supplier offerings.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- What Is Your AI Agent Buying? Evaluation, Biases, Model Dependence, & Emerging Implications for Agentic E-Commerce
- AI Agents Want to Shop for You: The Future of Agentic Commerce
- Vertical tacit collusion in AI-mediated markets
- The Agent Economy: Who's Most at Risk?
- The Coasean Singularity? Demand, Supply, and Market Design with AI Agents
- Remapping travel with agentic AI
- Agentic Commerce and Incentives
- Whom Do AI Agents Work For? Role Assignment Induces Sponsorship Bias in LLM Recommenders