Design for More Than the Super-User: Kelly Ommundsen on Who Hospitality Tech Leaves Behind

At the EHL HumanX Summit in Lausanne, Simone Puorto sat down with Kelly Ommundsen, Head of Digital Inclusion and a member of the executive committee at the World Economic Forum. Much of her work is with governments on digital transformation, which turns out to rhyme closely with hospitality. Both handle some of the most sensitive data there is, both are under pressure to modernise, and both keep reaching for new technology before they have asked what it is for. Across the conversation she kept coming back to two questions: who are we leaving out, and what problem are we trying to solve. The full conversation is available to watch below.

Bigger than the app

When hospitality talks about digital transformation, Ommundsen said, it usually means the guest-facing layer: the new loyalty app, the new kiosk, the new way a guest meets technology. That is the visible part, and it is the smallest part. The real change runs through the whole ecosystem, including the teams working behind the scenes to look after the guest.

Her advice is to treat it as a change-management problem rather than a purchase. She framed AI as a new coworker, one who arrives with useful skills but also upends the current ways of working, so the task is to bring everyone along. The technology should empower the workforce, she argued, not simply shrink it.

Design for more than the super-user

A second blind spot is who we design for. Teams tend to build for the super-user, the tech-savvy person who downloads the app and works the kiosk without thinking. Ommundsen made it personal. Her father does not own a smartphone, so a new app would leave him lost. Her mother has Parkinson's, and a check-in kiosk she cannot read or operate could mean she never reaches her room.

In hospitality, she pointed out, the user is anybody who walks through the door, from a grandmother to a five-year-old to a tech-savvy twenty-year-old. Simone added that the same guest has different needs at different moments, almost a different persona depending on the day, and she agreed. The job is to design across the whole spectrum of comfort with technology, not only for the confident end of it.

Peak hype, and a branding problem

On where the hype sits, Ommundsen was direct. AI is at the peak of its hype cycle. Part of the trouble is semantic. There is no shared definition, so everyone hears something different in the word and loads their own hopes and fears into it. The ground is also still moving, from large language models to what the field now calls agentic AI.

Her own line is firm. AI should never replace humans, full stop. It can do plenty of things better than people, she accepts, including forecasting, modelling, supply-chain optimisation and database management. The opportunity is to hand those tasks over so that people are freed for the work only humans do well. She would happily lose the seventy-five percent of her time that goes on email and informational meetings if it bought more room for creativity and human interaction.

She made the point with the interview itself. She could have been talking to an AI, she said, but it would not carry the spark and the uniqueness of a human exchange. That, for her, is what the summit was built to protect: excitement about the innovation, with the human kept at the centre.

Regulating something that won't sit still

Simone raised the idea of a regulated ratio between human and digital workers. Ommundsen was not keen on any fixed or universal number. What she does insist on is recognising the disruption AI will bring to the workforce, the way every major technology has. Cars, she noted, did away with the stable hands, the farriers and the drivers of the horse and buggy, and few would say the trade was a bad one. The lesson is not to fear the technology, but to put the right guardrails around it.

Those guardrails are hard to build, because the thing being regulated will not hold still. The EU AI Act is what much of the world is watching, but the deeper problem is that these systems are not static. They learn and adapt after they are deployed, and a new model with radical new capabilities appears every few months. Rules written for the state of play today date quickly.

This is where she would keep humans firmly in the loop. Take the example of firing most of your coders and leaving the work to an AI. When the code breaks, she said, you still need people who understand what the system did in order to fix it. You need audit and auditability, and a way to follow what the technology is doing. The aim is to unleash the innovation while protecting the values a society wants to keep, which she expects to be one of the defining challenges of the coming decade.

The mismatch behind her work

That speed gap is the basis of one of her initiatives at the Forum, the Global Regulatory Innovation Platform. Its starting thesis is a fundamental mismatch between the pace of rulemaking and the pace of technology. A piece of legislation might take two years, and that is generous, while a significant new model can arrive every three to six months. The platform looks at how to give regulators the tools to keep up.

There is an economic edge to this. Innovators and entrepreneurs move to where the rules are favourable, so a country that gets regulation wrong can lose the innovation along with the protection. There is also an information gap. Most policymakers have a surface understanding of how these systems work, not the deep technical grasp the engineering demands. Closing that gap, and turning it into a more collaborative conversation, is much of what her team is trying to do. It is a question for every industry that touches technology, she noted, which is to say nearly all of them, and the data-heavy ones like hospitality most of all.

What problem are we solving?

The interview closed on the HumanX theme, technology and humanity as collaboration rather than contest. Ommundsen brought it back to her own field. Governments, like hotels, hold the most sensitive data people have, their social security numbers, passports, health and education records, and both are under pressure to be more efficient and more centred on the person at the end. AI can help with that.

The discipline, she said, is to stay outcome-focused. Too often a project starts because a minister, or an executive, has decreed that AI must be implemented. The better starting point is the problem. If a guest's air conditioning fails and the room becomes unbearable at three in the morning, AI might diagnose the fault at the system level, but it will not fix the unit, and it will not soothe the person standing upset in the heat. Technology is a tool in the toolbox. The question worth asking, and asking again, is what problem we are trying to solve, not which new tool we happen to be holding.