Big 5 hotel chains all have AI assistants now. Independents have nothing.

I spent Tuesday morning watching the Hilton AI Planner announcement land with the calendar open on my second monitor. Accor in Q3 2025. Hyatt in Q4 2025. IHG in Q4 2025. Marriott in Q1 2026. Now Hilton. Five chains, six months, one product category — conversational AI search-and-concierge inside the brand app. The Big Five are complete. The category is officially table stakes at the chain level. There is nothing equivalent at the independent end of the market, and the gap between the two sides is now wide enough to see from the lobby.

The contrarian read — and I will defend it for the rest of this column — is that the gap is not a moat for the chain. It is a moat for the brand-affiliated property, which is a different thing. And the right response from the independent side is not to build. It is to outsource.

The six-month sweep, in calendar order

Walk the prints in sequence and the pattern is hard to miss.

Accor went first with the AI Travel Companion in Q3 2025 — a natural-language layer over destination discovery and ALL loyalty inventory, launched into the European leisure cycle. Hyatt followed in Q4 2025 with the AI Concierge, branded inside the World of Hyatt app and tuned for the resort and lifestyle book. IHG shipped Smart Planner in the same quarter, leaning harder into business-traveller use cases and route optimisation across the brand portfolio. Marriott rolled out its broader AI Assistant work in Q1 2026 — the RENAI search agent and the surrounding stack we pulled apart in a Marriott case study. And this morning, Hilton closed the set.

Five chains. One product shape: a conversational interface that takes “I want a hotel in Lisbon next weekend, walkable to dinner, under €300” and returns a curated brand-inventory result with loyalty stitched in. The product surface varies — search-first, concierge-first, planning-first — but the category is locked. (Interpretation flag: I am calling these five products one category. The vendors would not. Each release frames the work as differentiated. From the operator and guest side, they are not.)

The gap is the headline. The size is the preview.

Here is the part I want to be careful with. The chain-versus-independent AI-adoption gap is the obvious column, but the hard numbers are not all on the table yet. Forthcoming penetration data — early reads suggest roughly 100 percent of major-chain flags now ship a conversational AI surface, against around three percent at independents — is the figure the trades will run with over the next two weeks. The annual-spend gap follows the same shape: chain AI budgets are in the $2.5 million-plus range per brand programme; independents are spending somewhere between $5,000 and $15,000, mostly on chatbot widgets and review-response tools. (Interpretation flag: those penetration and spend figures are previews from forthcoming category coverage, not audited disclosures. Treat them as directional. I expect them to firm up by April.)

What is not preview is the category-completion fact. As of Tuesday morning, every Big Five chain has a shipping AI assistant in its consumer app. That part is on the record, and the Hilton release itself is the cleanest single citation for where the category lands today. For the audit-level read on the Hilton product specifically, the companion piece our team published this morning walks the feature surface.

Why the gap is the wrong moat to envy

The conventional read is that the chains have built a structural advantage and the independents are now permanently behind. I think that read is exactly backwards.

A $2.5M-plus annual AI programme at a chain funds a product that lives inside the brand app and routes guests to brand inventory. The moat accrues to the brand, not to the individual property. The Marriott AI Assistant is good for the Marriott flag in Lisbon — and it is also good for every other Marriott flag in Lisbon competing with that one. The chain product is a guest-acquisition layer for the chain. It is a guest-redirection layer between brand-affiliated properties.

The independent hotel does not have that product. It also does not have that internal competition. The right comparison is not “independent versus chain.” It is “independent versus the brand-affiliated property next door, which now has a corporate AI layer pulling guests through the brand funnel.” That comparison is real and it is getting harder for the independent every quarter the chains iterate.

The wrong response from the independent side is to build. A $5K–$15K annual budget does not fund a competitive conversational search product. It funds a worse one. The right response is to ride the PMS-and-distribution aggregators — Mews, Cloudbeds, Stayntouch — which are quietly building the AI surface that an independent cannot build alone and rolling it into the platform fee. The aggregator AI is the independent’s answer to the chain AI, and it is the only answer with workable unit economics.

What I am watching before April

Three threads.

Whether Mews, Cloudbeds, or Stayntouch ships a guest-facing conversational layer inside the next two quarters — and which one ships it as a default versus a paid module. The default-versus-paid choice is the one that decides whether the independent side closes the gap or settles into a permanent two-tier market.

Whether the chain AI products start ranking cross-brand inventory the way booking engines do. Today they do not. The day Marriott’s assistant surfaces a Hyatt result is the day the category becomes a meta-search business and the moat shifts again.

Whether the vendor-supplied voice and concierge stack — the Voice Agent Maturity Curve framework Eitan and I have been mapping for hospitality — starts converging with the brand-app AI assistants. The two product surfaces are still separate today. They will not stay separate.

Five chains, one category, six months. The independent gap is real. The moat is not where the headlines are putting it.

— Naomi covers hotel F&B and operator tech for TableTransfers. Tips: [email protected].