AI With Your Team, Not On Your Team: The First Rules Are Here

91% of hotel chains already use AI. Only 13% can show a measurable return. And the barrier they name first isn't the technology — it's people: 56% say they lack the skills.

That gap is where this week's news lives. While hotels were busy adopting, the rules started arriving.

Three signals in one week

• The first hospitality AI governance framework. The AI Hospitality Alliance published it for free: 5 pillars, 26 capabilities, and a "minimum viable" version for smaller operators. One pillar is entirely about people — guests and workforce.
• California's "No Robo Bosses" law. From July 2027, an employer can't rely primarily on an automated system to discipline or fire someone. A human must check the evidence, and the employee must be told in plain language that a system was involved.
• The workers' side. A University of Westminster professor writing in PhocusWire: algorithms already shape shifts and evaluations in tourism, usually without asking the people affected. A 2026 study links digital management tools to emotional exhaustion and lower performance among hotel staff.

Put together, the message is simple. AI that's imposed on a team underperforms. AI that a team chose to use pays off.

Why the adoption numbers don't show up in the P&L

Look at how hotels actually use AI today: 60% of daily use is someone opening a general chatbot on their own. No shared rules, no owner, no measure. Only 28% of chains have a company-wide strategy.

That's not a technology problem. It's a house-rules problem.

And the executives know it. At Destination AI in Washington, big operators talked about taking tasks off the GM's desk. Marriott and Cloudbeds were clear that cutting jobs isn't the goal. 59% of chains say the real win is freeing staff for higher-value work. That's the wingman model, written into survey answers.

Three house rules before you switch anything on

You don't need a 26-point framework for a 30-room hotel. You need three sentences everyone on the team can repeat:

• What the machine decides alone. Opening hours, parking, the late check-in code, a booking confirmation. Fast, repetitive, 24/7.
• What a human always decides. Complaints, refunds, exceptions, anything about a colleague's job. The machine can prepare it; a person signs it.
• How the guest knows. In Europe, the AI Act's transparency rules already require people to be told when they're talking to an AI. Say it openly — guests don't mind a fast machine, they mind being fooled.

Then one question to the front desk: "Which task do you hate most?" Start there. Adoption follows relief, not orders.

The takeaway: this week, write your three house rules on one page and ask your team which task they'd hand to a machine first. AI is my Wingman — but a wingman only works if the pilot chose it. The machine takes the repetitive load; your people keep the welcome, the judgment and the last word.