Google, OpenAI & Booking.com's Opposite AI Bets

The industry's biggest players are looking at the same AI revolution and arriving at fundamentally different conclusions about where the value sits.

Not minor variations in approach, but opposite philosophies about what travellers actually want from technology. I think the divergence itself is the story, and the companies that win won't be the ones who picked the right side. They'll be the ones who stopped picking.

Forget the individual strategies for a moment. The fact that billion-dollar companies are placing contradictory bets tells us we're in a genuinely uncertain period. And if you're running a hotel or a travel business, understanding these competing visions matters more than adopting any single piece of technology.

Google Wants the Whole Funnel

Google is moving aggressively to collapse the travel planning journey into a single AI-powered experience. Their recent Think with Google piece frames it explicitly. AI Overviews and AI Mode aren't just surfacing links anymore. They're "reasoning." They're solving what Google calls the "complexity tax" of travel planning.

The pitch to travel brands is clear. Search is moving from information to intelligence. Google wants to connect "inspiration to inventory," taking you from a vague idea like "family beach holiday, Southern Europe, September" all the way through to a bookable result without ever leaving the ecosystem. In their own words, there's a critical distinction between "chatting" and "booking," and Google believes it can bridge that gap better than anyone.

My read is that they're positioning to own every stage of the journey, from the first daydream to the confirmation email. If they pull it off, every intermediary in the chain becomes optional.

OpenAI Tried It and Backed Away

Here's where it gets interesting. OpenAI tried the same thing and retreated.

They'd been testing integrated checkout features inside ChatGPT, letting users book hotels without leaving the chatbot. Now they're pushing transactions into third-party apps instead. TD Cowen analysts called it a "stunning admission", and Booking.com and Expedia stock surged on the news.

The reality they hit is one that anyone in travel distribution already knows. Handling the messy parts of commerce, the payments, cancellations, refunds, customer service complaints, and legal liability, is operationally complex and incredibly difficult to scale. Skift's coverage noted the sheer difficulty of aggregating all that inventory and making it bookable across a single platform.

The precedent nobody mentions

In the 2010s, Google and several metasearch players tried to insert themselves directly into the booking flow. Most pulled back. The operational weight of travel transactions has a way of humbling even the biggest technology companies. OpenAI appears to have learned that lesson faster than most.

Booking.com Is Betting on Humans

And then there's what I find the most counterintuitive move of all. At the Qualtrics X4 conference last week, Booking.com positioned human support as its key competitive differentiator. Not AI. Humans. Their approach uses AI to handle triage, identifying intent and providing agents with relevant context, but the company sees human connection as the thing that will set them apart from competitors.

The financial picture makes this even more compelling. Booking.com's AI investments have already driven customer service costs down 10% per booking, fuelling a $700 million growth reinvestment programme for 2026. They're using AI savings to invest in better human service.

If you follow that line of reasoning, it starts to look less like a choice between humans and AI and more like a deliberate sequence. Automate the commodity layer first, then reinvest in the moments that actually need a person. That's not a contradiction. It's a strategy.

"Rather than humans versus AI being the defining strategic choice, it becomes humans because of AI."

The Single-Strategy Trap

My view is that all three are partially right, and none of them are fully right if they commit to a single approach exclusively. Travel decision-making is deeply personal, and the same traveller who books a weekend city break on autopilot might spend three weeks agonising over where to take their parents for a milestone anniversary. The trip determines the channel, not the other way around. No single interface handles that range without losing people at the edges.

If you force every customer down one path, whether that's a chatbot or a phone line, you'll win some and alienate others. The companies that offer genuine choice will capture the widest share of the market. Not because choice is a nice-to-have, but because travel is too varied and too personal for a one-size-fits-all approach.

Why boutique travel agents still win

This is why bespoke travel agencies continue to thrive despite the platforms having every structural advantage. With over three million properties on Booking.com alone, AI can surface dozens of plausible options for any given traveller. But more choice can be paralysing. The value of a great travel agent is curation. The personal relationship. Knowing the client well enough to cut through the noise. That's a proven model, and it's exactly what Booking.com appears to be trying to replicate at scale.

The Bigger Shift Nobody's Talking About

If agentic commerce continues to grow, with AI agents shopping rates and availability via API rather than humans browsing websites, the traditional OTA front-end becomes less central to the equation. Booking.com already distributes availability through resellers and API connections. Perhaps the future isn't Booking.com as a destination you visit, but Booking.com as an enabler whose inventory and intelligence powers the wider ecosystem.

If that happens, it frees up even more resource for the human element. The agentic layer handles the transactional commodity work. The humans handle the high-value advisory moments.

"The best companies won't be the ones that picked the right side. They'll be the ones that understood there were no sides to pick."

I think that's where this is heading. Not a war between technology and service, but a rebalancing. A new division of labour between machines that are extraordinary at breadth and humans that are irreplaceable at depth. For hotels, for travel agents, for anyone in this industry, the question isn't whether to adopt AI. It's where you deploy humans and where you let machines do what they're better at.