AI for Hotel HR: 5 Takeaways from Cornell HR in Hospitality 2026

The big promise of AI in hospitality? Give managers their time back without taking the humanity out of the work.

That was the through-line at "Leveraging A.I. to Improve Well-Being, Engagement and Performance," a panel at Cornell HR in Hospitality 2026. Three founders working at the intersection of AI and hospitality talked through where the technology is actually helping, where it creates risk, and how to tell a real tool from a good sales pitch. The panel featured Zev Eigen CEO of Fourth Party AI, Luke Fryer CEO of Harri, and Marissa Fetter Hochster CEO of Hylite, moderated by Professor Mike Maffie.

The HR tech market is crowded with AI tools, and many of the vendors selling them can't clearly explain what their product actually does. Meanwhile the workforce is (understandably) wary of being "optimized." Get the tech stack right and you free your managers to lead. Get it wrong and you spend a fortune automating the wrong things, or worse, automate away the parts that make hospitality, well, hospitality.

The shared conviction from the panel: the point of AI isn't to remove the human. It's to give them room to actually be one.

1. The opportunity isn't replacement. It's better hospitality, concentrated where it counts.

Zev Eigen pushed back on the idea that AI is going to either save everything or destroy everything. His view was more nuanced: AI will likely create more separation between low-touch and high-touch hospitality experiences. Automated check-in works for some guests and some brands. But the more transactions technology handles, the more valuable real human service becomes. If a kiosk takes care of the basics, your team needs to be even better at the moments that require warmth, judgment, empathy, and conversation.

Luke Fryer, CEO of Harri

Hospitality, Luke Fryer argued, doesn't face the same kind of foundational disruption as industries where removing the human is the whole cost-saving play. People will still eat out. People will still sleep away from home. Service will still matter. The more immediate opportunity is helping frontline managers spend less time on admin and more time leading people. Scheduling, timesheets, onboarding, HR actions, compliance workflows, employee communication — all of it is fair game.

The question for HR leaders isn't "how do we replace managers?" It's "how do we make the manager's job easier?" That's a much better question.

2. Start with the problem. Then measure whether AI actually solved it.

One of the clearest warnings from the panel: don't implement AI just because someone says the company needs "an AI plan."

As Luke put it, the project shouldn't be AI itself. The project should be solving a real business problem. Where are managers spending too much time? Where are employees getting stuck? Where are candidates dropping out? Where is compliance risk building? Where is turnover happening fastest? Start there. Then ask whether AI can help.

Once you've implemented something, measure it honestly. Is hiring actually faster? Is retention improving? Are managers actually spending less time on admin? Run a pilot. Compare before and after. Track whether adoption increases over time, or whether the tool quietly becomes shelfware.

Too many companies, Zev noted, spend millions on technology and then measure success by whether people "like it." That's not enough.

3. Don't let vendors hand-wave.

This was one of the most useful, and funniest, pieces of advice from the session.

Vendors love vague language. "Proprietary models." "Advanced AI." Zev warned HR leaders not to let any of it pass without an explanation of what the tool actually does. Some products, he noted, are essentially expensive wrappers around the same large language models you can get yourself for $20 a month.

His advice: make vendors explain the technology in plain English. What's proprietary? What data is being used? What model is powering it? How is it tested? What problem does it solve?

If the salesperson can't explain it clearly, STAY AWAY.

4. AI can spot retention risks before people leave.

Short-cycle turnover was a major focus for Luke, especially employees leaving in the first 90 days. In restaurants and hotels, the cost of early turnover is enormous, and the drivers are often surprisingly practical: poor training, schedule dissatisfaction, lack of recognition, or simply not having the right tools to do the job.

AI can help spot the patterns earlier. If new hires aren't completing training, are unhappy with schedules, or are quietly disengaging, managers can step in before the exit interview.

The goal isn't to surveil people. The goal is to notice problems while there's still time to fix them.

5. Recognition may turn out to be AI's most human application.

Marissa Fetter Hochster told the panel about her favorite Starbucks barista, Darrell. She went out of her way to that specific Starbucks because of him, and his manager had no idea. The praise lived in Marissa's head and nowhere else.

Marissa Fetter Hochster, CEO of Hylite

That story captured something hospitality leaders know instinctively: guests often remember the person more than the product. The problem is that recognition usually disappears into a nice memory or a casual thank-you at checkout. It rarely reaches the employee, the manager, or the organization.

Hylite is built to change that. AI-enabled recognition tools can capture guest praise in real time, surface standout employees, identify patterns, and even turn positive guest feedback into reviews and marketing content. More importantly, recognition supports retention by helping employees feel seen. Who doesn't love that?

The key to successful AI adoption

The heart of this AI conversation wasn't really about computers, algorithms, or shiny new tools. It was about trust, judgment, and time.

AI can help hospitality companies hire better, schedule smarter, recognize employees faster, and spot retention risks earlier. But only if HR leaders ask the right questions before they buy the tech.

What problem are we solving? How will we measure success? Could this create bias or compliance risk? Are employees actually using it? Does it make the manager's job easier? Does it protect, or erode, the human touch?

Because in hospitality, the human touch isn't what AI replaces. It's what AI is supposed to protect.