Hospitality has an AI blind spot

There’s a version of the AI story in hospitality that sounds reassuring.Efficiency gains. Smarter forecasting. Better staffing. Less admin. And much of it is true. But over the past few months, while working on a broader project on AI and ethics in business, I’ve become increasingly convinced that we’re missing the more important risk - particularly in foodservice. It’s not about chatbots or customer-facing gimmicks. It’s what’s happening quietly in the back office.

More and more operators are relying on systems that don’t wait to be asked. They forecast demand, build rotas, trigger orders and optimise decisions automatically. And they often do it well enough to avoid immediate scrutiny.

But here’s the problem: Most businesses were not doing structured data analysis before. Now they’re moving straight to automated decision-making.

At the same time, the human capacity to question those decisions is shrinking. The skills that once allowed operators to spot when something didn’t feel right are being eroded - because the system has taken over the work that built them.

The result is a growing gap between trusting AI and understanding it.  And in that gap sits a risk that doesn’t announce itself. It doesn’t fail loudly. It fails quietly, behind dashboards that look precise, polished and reassuring.

In this week’s issue, I explore:

• Why we instinctively trust AI output more than we should

• How “agentic” systems change the nature of risk entirely

• Why smaller operators may be the most exposed

• And the uncomfortable question the sector isn’t yet asking

• The question isn’t whether AI will transform hospitality. It’s whether anyone will notice when it does it badly.

Read the full article