h2c, the hospitality research and consulting firm, has published a new study concluding that artificial intelligence adoption is now widespread among hotel chains, but that enterprise-level readiness for the technology remains limited.

The report, distributed via Hospitality Net, frames a tension that hoteliers recognize across their operating P&Ls: AI experimentation is multiplying on property, yet the plumbing — data, governance, integration — required to scale pilots into firm-wide capability has not kept pace.

What's the gap between adoption and readiness?

For operators, the two findings are not contradictory. A single property can run a conversational booking widget, an automated revenue-management alert, or a back-of-house scheduling assistant and still be counted among the "adopters" of AI.

Enterprise readiness, by contrast, requires those tools to share a connected data layer across property-management, central-reservation, revenue-management and point-of-sale systems. It also requires that ownership governs outputs with the same discipline applied to labor scheduling or purchasing.

The practical implication: any return on AI spend currently accrues at the unit level, not at the portfolio level.

Why does this matter for hotel operators now?

Three operating pressures sit at the center of the AI conversation inside hotel companies today:

• Labor costs: wage inflation and service-hour caps continue to compress margins, and vendors are selling scheduling, forecasting and guest-service tools as relief valves
• Distribution economics: a fragmented stack across brand.com, OTA and GDS channels makes personalization hard to execute; AI is pitched as the bridge
• Guest expectations: personalization at check-in is increasingly table stakes, and operators without connected data fall behind regardless of how many pilots they run

The h2c finding suggests adoption is outrunning integration, leaving many hotel groups with a portfolio of disconnected tools rather than an enterprise capability.

What changes next?

For ownership groups and brand operators, the test will not be which AI tools they buy but whether they rebuild the underlying data and governance architecture fast enough to capture scale returns. Vendors that can integrate cleanly with PMS, CRS and labor systems — rather than sell standalone point solutions — will likely see a procurement advantage as the enterprise-readiness gap narrows.

As multi-property groups move into the next budgeting cycle, expect AI line items to face the same scrutiny as labor percentage, cost of goods and distribution margin — and expect the operators with integrated data layers to post the cleanest returns.