Davos's AI Conversation Was About Everything Except Hospitality

I spent Tuesday night doing something a little strange. I pulled the public WEF program for “Collaboration for the Intelligent Age” — the 2025 Davos theme — and I scrolled through every panel with “AI” in the title looking for a name I recognized from our beat. A restaurant CEO. A hotel CEO. A QSR CFO. A loyalty head. Anyone who, on a normal earnings call, would be the one explaining how AI is going to reshape ordering, scheduling, or guest experience.

I scrolled for a long time. I didn’t find them.

That is the story I want to tell this week, on the third day of the World Economic Forum’s annual meeting in Davos. The AI conversation at the top of the global agenda is loud, expensive, and confident — and hospitality, the industry that touches more daily human transactions than almost any other, is barely in the room. My read: absence is the story. The largest operators in our category are deliberately quiet on AI in public, and that posture is itself a data point worth reading carefully.

The panels were about infrastructure and labor — not about your line

If you skim the CNBC recap of the Davos AI conversation (cnbc.com), the contours are clear. The CEOs who showed up to talk about AI are talking about two things: the cost and shape of the infrastructure layer (chips, power, data centers, capital intensity), and the second-order labor question (which jobs get reshaped, which get displaced, who pays for retraining). The voices being quoted — Mendelsohn on the enterprise rollout pace, Mensch on European model competitiveness, Summers on the macro implications — are paraphrased there, and I’m not going to put words in their mouths beyond that. Go read the recap if you want the texture.

McKinsey’s own Davos debrief (mckinsey.com) reinforces the framing: the consultancy-class read of the week is that 2025 is the year AI conversations move from “what is it” to “what does it cost to deploy at scale and where does the workforce land.” The WEF’s own writeup of AI leaders at the meeting (weforum.org) leans similarly — governance, frontier safety, sovereign compute.

Notice the through-line. Every one of those framings is upstream of hospitality. None of them is about the unit economics of a 30-second drive-thru, the loyalty CRM, the back-of-house labor split between cooks and front-of-house service, the franchisee P&L. Those are the questions our category is actually living with — and they’re being discussed in trade press, on earnings calls, and in vendor decks, but they’re not on the Davos main stage.

Why the silence — and what I think it actually means

The easy explanation is that Davos is a CEO stage and hospitality CEOs prioritize their own industry conferences. Fair. But it’s incomplete. The CEOs of consumer banks, pharma, and energy companies — also industries with their own conference circuits — did show up to AI panels this week. Hospitality’s biggest names did not headline them. McDonald’s, Yum, Marriott, Hilton, Compass — the operators whose Q4 2024 earnings will land in the next few weeks — were not the faces of this conversation.

My read: the silence is strategic, and it’s coming from three places at once.

First, the failed pilots are still fresh. The IBM partnership winding down at McDonald’s last summer is the most public example, but the pattern of “automated order-taking is harder than the slide deck implied” is widely understood inside the operator class. CEOs who got out over their skis on AI in 2023 are deliberately not doing it again on a Davos stage in 2025.

Second, the labor politics are unusually hot. Saying “AI will reshape our workforce” at Davos is a different sentence when your workforce is 2 million hourly employees and your franchisee community reads everything you say. The hotel side has the same dynamic with housekeeping and front-desk roles. There is no upside to being the CEO who got quoted on labor displacement during the week the WEF spent talking about labor displacement.

Third — and this is the one I find most interesting — the actual AI rollouts in our category are happening at the vendor and infrastructure layer, not the brand layer. The Toast, Olo, Square, PAR, Oracle Hospitality side of the stack is where the model integrations are landing. Brand CEOs can plausibly let the platforms do the talking on AI for another quarter or two, and several of them clearly chose to. As we’ll explore in a forthcoming Mise piece on hospitality-AI economics, the economics of who captures the AI margin in hospitality are genuinely unsettled, and the brands have reasons to wait.

What operators should actually take from Davos this week

If you run a restaurant group or a hotel portfolio and you watched any of the Davos coverage, here is what I’d suggest you actually do with it.

Read the infrastructure conversation for what it tells you about cost. The reason CEOs in other industries are talking about compute, power, and capital intensity is because the unit economics of running AI workloads are not falling as fast as the headlines suggest. If you are budgeting 2025 AI spend on the assumption that inference cost is collapsing, the Davos signal is to recheck that assumption with your vendors. The capex story your POS provider is telling you has to come from somewhere.

Read the labor conversation for what it tells you about timing. The Davos consensus is that labor reshaping from AI is real but slower and messier than the 2023 framing implied. For hospitality that maps to a specific operational truth: voice ordering, AI scheduling, and back-of-house automation are not arriving as a single wave in 2025. They’re arriving as a four-to-six-quarter staircase, and as our later framework on the voice-agent maturity curve argues, the operators who win will be the ones who pace their rollouts against that staircase rather than chasing the keynote.

And read the absence for what it tells you about your peers. The largest brands in our category did not use this week to plant a flag. They are watching the EU AI Act compliance deadlines coming on February 2, watching the Q4 earnings cycle, and watching each other. If you are a mid-market operator and you’ve been feeling behind because you don’t have a public AI narrative — Davos this week is evidence you are not as behind as you thought.

The AI conversation will get to hospitality. It just didn’t this week, on purpose. That’s worth noting now, so we can mark the moment it changes.

— Hana edits The Pass. Tips: [email protected].