Public data on AI in hotels is sparse, because official surveys report AI use for whole economies, not for lodging alone. Statistics Canada found that 19.2 percent of Canadian businesses used AI in the 12 months before its spring 2026 survey, and the U.S. Bureau of Labor Statistics projects hotel desk clerk employment to grow 2 percent from 2025 to 2035. Controlled research on guest-facing chatbots points to a narrower role for them than marketing suggests.
Key takeaways
- Statistics Canada reported 19.2 percent of businesses using AI in the Canadian Survey on Business Conditions for the second quarter of 2026, up from 12.2 percent a year earlier, with no figure published for accommodation.
- Among AI-using businesses, 28.2 percent used chatbot-style tools, and security or privacy worries were the biggest barrier at 13.4 percent.
- BLS projections via O*NET show hotel, motel, and resort desk clerks at 264,600 jobs in 2025 and 269,500 in 2035, a rise of 2 percent.
- Two experiments in a hospitality journal found that chatbot service recovery produced lower satisfaction than human recovery.
- The FTC's 2024 Marriott and Starwood order shows what is at stake when guest data is stored: more than 344 million customers were affected by three breaches.
What national data say about AI use, and what they leave out
The Canadian Survey on Business Conditions asked each business whether AI had been used in its goods or services over the previous 12 months. The second-quarter 2026 results, released June 11, 2026, came from 9,251 responses collected between April 1 and May 6, 2026. Overall use stood at 19.2 percent, against 12.2 percent in the second quarter of 2025 and 6.1 percent in 2024. Another 40.0 percent of businesses said AI was not relevant to them.
Highest use was in information and cultural industries at 42.3 percent, finance and insurance at 40.4 percent, and professional, scientific, and technical firms at 32.4 percent. The Statistics Canada analysis does not report accommodation and food services, so no verified Canadian hotel figure exists in that release. A U.S. equivalent also could not be verified here, because the Census Bureau pages for its business survey could not be opened.
Chatbots, privacy, and training: what AI-using firms report
Among Canadian businesses that use AI, data analytics was the application named most often (36.6 percent), ahead of text analytics (34.5 percent) and chatbot-style agents (28.2 percent). Chatbot use rises with firm size: 50.5 percent of the largest AI-using firms reported virtual agents. Barriers matter too. Security and privacy worries topped the list at 13.4 percent, and cost came next at 10.6 percent.
On jobs, the agency says plainly that the employment effect of AI is still unclear. What it did measure is organizational change: 44.4 percent of AI-using businesses changed training or staffing practices, and 32.0 percent of those trained existing employees. These are all-industry numbers, so they describe the general pattern and not how a hotel group operates.
Front-desk jobs: a projection, not a forecast of AI effects
Self-service check-in and digital keys are often cited as a threat to front-desk work, so the official projection is worth reading carefully. O*NET OnLine, which publishes Bureau of Labor Statistics projections, shows 264,600 hotel, motel, and resort desk clerks in 2025 and a projected 269,500 in 2035. That is growth of 2 percent, labeled slower than average, with 35,500 projected annual openings. The page was last updated October 6, 2026.
A projection of modest growth does not isolate AI. It reflects demand for lodging, turnover, and technology together. It also does not say that individual tasks, such as answering routine questions, will not shift to software. Task exposure and job loss are different measures.
Guest service recovery: what experiments found
When a stay goes wrong, hotels increasingly route complaints to chatbots. A 2023 paper in the Journal of Hospitality and Tourism Technology (volume 14, issue 3, pages 416 to 429) by Yimin Zhu, Jiemin Zhang, and Jifei Wu ran two experiments on this. Chatbot recovery produced lower customer satisfaction and lower revisit intention than recovery by a human employee. The gap was larger when the recovery was symbolic, such as an apology, than when it was economic, such as a refund. Immediate recovery narrowed the gap.
The abstract available for this article did not report sample sizes, and the chatbots studied predate current generative models. The finding is a dated experimental result, not a ruling on today's tools.
Guest data: the cost of storing what AI systems feed on
Personalization and chat tools run on guest records, which makes data security a design requirement. On December 20, 2024, the Federal Trade Commission finalized an order against Marriott and Starwood over three breaches that exposed information of more than 344 million customers, including passport information, payment card numbers, and loyalty numbers. The FTC announcement lists the terms: a comprehensive information security program, a retention limit, a deletion request link for U.S. customers, and restoration of stolen loyalty points. The case did not involve AI, but any system that ingests guest data inherits the same obligations.
What each source measures
| Source | What it measures | What it does not show |
|---|---|---|
| Statistics Canada, Q2 2026 | Share of all businesses using AI, uses, barriers | Hotels specifically, or job outcomes |
| BLS projections via O*NET | Desk clerk employment, 2025 to 2035 | The share of change due to AI |
| Zhu, Zhang, and Wu, 2023 | Guest reaction to chatbot recovery | Current generative AI performance |
| FTC order, 2024 | Guest data security duties after breaches | AI-specific rules |
FAQ
Are hotels widely using AI?
Industry-specific official figures were not found. The Canadian all-industry figure is 19.2 percent, and hotel use cannot be read from it.
Will AI eliminate hotel front-desk jobs?
The official projection shows slight growth to 2035, but it does not model AI separately. Treat claims of mass replacement as unsupported by that data.
Do guests prefer chatbots for complaints?
In the 2023 experiments, chatbot recovery scored lower than human recovery on satisfaction and revisit intention, especially for symbolic recovery such as an apology.
What laws apply to guest data used by AI tools?
That depends on the jurisdiction and the company. The FTC order illustrates one federal enforcement route in the United States. This article is general information, not legal advice, and it does not address health matters or individual circumstances.
Related reading
- AI in large retail operations
- Solving complex problems with AI
- AI chatbot experiences built on GPT
- AI in learning environments
More in AI Blog.
Last verified: October 2026
Sources
- Analysis on artificial intelligence use by businesses in Canada, second quarter of 2026, Statistics Canada, June 11, 2026
- Hotel, Motel, and Resort Desk Clerks national employment trends, O*NET OnLine with U.S. Bureau of Labor Statistics 2025-2035 projections, 2026
- Who did what and when? The effect of chatbots' service recovery on customer satisfaction and revisit intention, Zhu Y, Zhang J, Wu J, Journal of Hospitality and Tourism Technology 14(3), 2023
- FTC Finalizes Order with Marriott and Starwood, Federal Trade Commission, December 20, 2024