Adriana Lee

Model Context Protocol is becoming a prerequisite for AI travel distribution. In the agent era, not being machine-readable will be the fastest way to become invisible.

TikTok inked a deal to sell a significant portion of its U.S. operation to American investors. The sale may bring clarity, but not comfort, as travel brands wait to see the effect on the algorithms driving discovery.

Any travel brand can now submit apps and get listed in its new app store — which means Booking.com and Expedia, early movers in OpenAI’s app ecosystem, will soon get new competition to own ChatGPT’s travel conversations.

United rewrote its mobile boarding, disruption handling, and personalization by weaving AI throughout its app — not just layering on a chatbot. The real question is whether the automations will ultimately help smooth travel or, by introducing errors or hallucinations, add to the travel friction.

Google’s new beta feature turns nearly any earbuds paired with Android phones into live AI translators. That makes language more accessible — and less of a friction point for travel.

Google is collapsing the gap between quick answers and deeper chats in mobile search. For travel brands, this could mean more queries — from inspiration to booking — will run through conversational interfaces, not blue links.

For a travel sector eager to automate, the basics still matter — especially continuity between bots and human travel agents. Selectour and Kleio tackle the handoff problem, and the effort could prepare advisors for the agentic future. 

As AI models like ChatGPT, Gemini, and Perplexity become the new travel gatekeepers, brands are rewriting SEO playbooks for an era where visibility hinges on being mentioned, not ranked.

Hotel tech investors are treating AI guest management as core infrastructure, not a side project. Duve’s new funding gives it room to test that thesis with big brands and new markets.

AI is reshaping travel’s workforce faster than executives can publicly acknowledge. The winners are efficiency and automation, and the losers, at least for now, are thousands of people.