AI in hospitality: Where should Singapore’s F&B operators draw the line?

If AI can reduce some workload, people can spend more time where they add the greatest value.

In July 2026, Enterprise Singapore and UOB launched the F&B artificial intelligence (AI) and Digital Integration Programme, aimed at supporting at least 200 F&B businesses over the next two years. Eligible businesses can receive advisory support and up to 50% funding support for selected AI-enabled solutions and integration services.

According to the Ministry of Manpower’s inaugural report on AI adoption amongst firms, 70.7% of firms that have adopted AI have already seen improvements in worker productivity.

Yet adoption within Singapore’s F&B sector remains relatively low, at just 21.9%.

At the same time, manpower pressures facing the sector are unlikely to disappear. At this year’s National Day Rally, Prime Minister Lawrence Wong reiterated that Singapore will continue to manage foreign manpower in a careful and controlled way. Instead, more of Singapore’s future growth will need to come from technology and productivity.

For hospitality operators, that makes the conversation around AI even more important. If simply adding more workers is not the long-term answer, we need to think more carefully about how technology can help existing teams operate more effectively.

But that brings us to the bigger question: not simply whether we should use AI, but where it belongs within the business, and where it does not.

To answer that, I think we first need to distinguish between hospitality and the business of hospitality.

Hospitality itself is about looking after people. It is reading a table, understanding what a guest needs, having the right conversation at the right moment, and ultimately creating a great experience.

The business of hospitality is everything that enables that experience: rostering, purchasing, forecasting, reporting, compliance, recruitment, and administration.

That distinction is important because I believe the greatest opportunity for AI today sits largely in the latter. I tend to think about that opportunity across three areas: time, information, and decisions.

The first is time. Research, reporting, communications, and data processing all take time. If AI can reduce some of that workload, our people can spend more time where they add the greatest value.

The second is information. A UOB survey conducted in June found that six in 10 F&B operators in Singapore still have most of their digital solutions operating as standalone systems, while about one in three has yet to adopt AI tools at all.

For many operators, sales, reservations, purchasing, labour, inventory, and marketing data still sit separately. AI creates an opportunity to connect those pieces and give operators a clearer view of the business.

And once you have better information, you can begin making better decisions. That, to me, is where AI becomes particularly interesting. Saving time may be AI’s biggest immediate benefit, but over the longer term, I believe decision-making represents the greater opportunity. AI can help management understand not only what happened, but why it happened, what might happen next, and what we may want to do about it.

The important word, however, is help. There are decisions I would never allow AI to make alone: staff performance issues, important guest relationships, major creative decisions, or sensitive HR matters.

AI can provide another perspective. It just does not get to vote. Human accountability needs to remain at the end of the process, particularly when decisions affect our people, our guests, or the experience we are trying to create.

For the same reason, I do not think the conversation around AI in hospitality should simply be about how many jobs it might replace. The more immediate question is how it changes the work people are already doing. The early numbers in Singapore point towards this more nuanced picture: amongst firms that have adopted AI, only 6.2% reported reductions in headcount, whilst 18.9% reported redesigning job roles, and 13.9% created new AI-related jobs.

My view is that AI will replace tasks before it replaces people. A typical hospitality role has repetitive or administrative tasks sitting around it, from reporting and rostering to research, purchasing, forecasting, and communications. If some of those tasks can be handled more efficiently, our people have greater capacity to focus on judgement, creativity and human interaction.

None of this means operators should rush to automate everything. High implementation costs and a lack of in-house expertise remain the two most commonly cited barriers to AI adoption amongst Singapore businesses, at 44.9% and 42.4%, respectively.

AI is also only as reliable as the information and instructions it receives. It is not a magic bullet, and introducing more technology does not automatically make a business smarter.

For that reason, I believe adoption should be progressive. Start by defining the problem. Use AI to assist with research and analysis. Use that information to support better decisions. As confidence in the technology and underlying data grows, businesses can explore deeper integration.

Over time, operators could bring sales, reservations, labour, purchasing, and cost information into one view, improving visibility and forecasting. From there comes simulation: what happens if reservations increase by 20%? What if a key ingredient increases 15% in price? What happens if staffing levels change?

Eventually, AI could help recommend the most appropriate response and even action certain low-risk decisions within parameters established by management.

But the principle should remain the same throughout that journey: people stay in control of the decisions that affect the guest, the team, and the brand.

Ultimately, the goal should never be AI for the sake of AI. It should be about creating a smarter and more efficient hospitality business without losing what made hospitality successful in the first place.

We should not be trying to automate hospitality. We should be automating the unnecessary friction around the people who deliver it.