AI architecture transforms travel and hotel distribution

Alkemy Iberia released a study on how "agentic" artificial intelligence is being deployed in the travel sector. The report argues that the most significant AI transformation occurs not at the user‑facing layer (chatbots, copilots) but in the intermediate "middle layer" that links models, data, enterprise systems and transactional execution. Fragmented architectures across hotels, airlines and booking platforms hinder consistent AI decision‑making, prompting a push for unified protocols such as MCP, A2A, UCP, AP2 and ACP, supported by major players like Google, OpenAI, Stripe and Microsoft. Experts stress that control of this middle layer determines who truly governs AI‑driven decisions in tourism.

A related analysis highlights how automation is reshaping hotel distribution. When a hotel’s inventory is offered at a net rate to a B2B operator, the contract often allows free cancellation up to a certain deadline. Automated systems monitor wholesale markets for lower net rates, cancel the original booking and re‑book at the cheaper price, leaving the hotel with reduced revenue while the end‑customer sees no change. This invisible arbitration stems from hyper‑connectivity, algorithmic margin optimization and wide‑open cancellation windows, creating new revenue‑management challenges for hoteliers.