Image: EA Hospitality Pulse (publisher-supplied) · Licensed by the publisher

Africa's hotels adopted AI first. Now comes the hard part.

By Onyango George (OG)

Hotel chains across Kenya, Rwanda and Tanzania have integrated artificial intelligence faster than any region on earth. The industry's own data suggests that was the easy part.

On the last day of March, in a conference hall in Nairobi, Jameel Verjee stood up in front of Africa's hospitality industry and announced that his hotel group would hand a piece of its guest relationship to a piece of software.

Verjee is the founder and chief executive of CityBlue Hotels, a chain that grew from a single property in 2013 into one of the continent's fastest-expanding hospitality brands, with hotels in Nairobi, Mombasa, Kigali, Dar es Salaam and Accra, and openings planned in Uganda, Ghana and Zambia. At the Future Hospitality Summit Africa, he introduced a partnership with Inntelo AI, a two-year-old London start-up whose "AI-native" platform is meant to run guest messaging, task routing and service coordination across a hotel in real time, not as a chatbot bolted onto a booking form but as something closer to a nervous system.

"As we scale, AI-native operations become essential to delivering consistency, speed and quality across multiple geographies," Verjee said. "Just as importantly, we are shaping how AI is applied within an African context." Asif Alidina, Inntelo's founder, framed CityBlue as proof of where the industry was heading: "a natural partner" for technology he described as "purpose-built for hospitality."

It was one announcement, on one continent, at one summit. But it fits a pattern that has quietly upended a piece of received wisdom about global hospitality: that African hotels are technology laggards, waiting for the rich world to perfect a tool before it trickles down. The data says the opposite.

A study published last October by the research firm h2c, commissioned by the Polish booking-technology company Profitroom, surveyed 171 hotel chains across four world regions. It found that 57 per cent of hotel businesses in the Middle East and Africa had already integrated AI-driven features into their operations, against 35 per cent globally, 30 per cent in Europe, 30 per cent in the Americas and 29 per cent in Asia-Pacific. MEA hoteliers also reported the highest trust in AI of any region, at 7.1 out of 10 against a global average of 6.6, the greatest comfort with AI-set pricing, and the lowest share worried that AI would damage the guest experience.

"African hotels are demonstrating remarkable leadership in turning AI potential into business reality," said Katarzyna Raiter-Łuksza, Profitroom's director of product. "What's particularly striking is not just the adoption rate, but the confidence African hoteliers have in this technology compared with their global counterparts."

If the story ended there, it would be a tidy piece of counter-programming: the emerging market that leapfrogged the incumbents. It does not end there, because the h2c study measures procurement, not proof. It counts hotels that bought or switched on an AI feature. It does not ask what that feature actually changed. A second, much larger study published this month asks the harder question: not who adopted AI, but what adopting it actually bought.

The State of Distribution 2026, now in its third year, is built by the hotel-technology group RateGain together with New York University's Jonathan M. Tisch Center of Hospitality and HEDNA, the trade body for hotel distribution executives. It draws on more than 270 hotel brands and 58,000 properties across 141 cities and 53 countries, covering the year to November 2025, which makes it one of the largest instruments measuring how hotels actually behave, rather than what they say they plan to do.

Its central finding: more than half of hotels worldwide now use or are procuring generative AI. Fewer than one in ten report that it has cut their manual workload by more than 30 per cent. More than 80 per cent of commercial teams still spend one to two days a week manually pulling reports out of systems that do not talk to each other, and fewer than 30 per cent have bought dedicated software to fix that. Only 8 per cent of hotel chains globally have a company-wide AI strategy at all.

"Buying AI is easy. Getting value from it is not, and this report shows most of the industry is still stuck between the two," said Bhanu Chopra, RateGain's founder and managing director. "The advantage will not go to the hotels with the most tools. It will go to the ones that turn their technology into better decisions and give their teams their time back. That shift has not happened yet, and it is the single biggest opportunity in front of the industry."

The mechanism behind the gap is not mysterious, and it echoes what researchers have found when AI has been pushed into other high-stakes, high-friction industries: the tool works when tested against clean data in a pilot, then meets a live property with a decade-old booking system, a front desk that changes staff every eighteen months and guest records scattered across three platforms that were never designed to share information. Vanja Bogicevic, who directs NYU's Hospitality Innovation Hub, put it as a widening distance between ambition and infrastructure. "AI is reshaping how travellers discover hotels, how commercial teams work, and how pricing and demand decisions are made," she said. "These new benchmarks show where hotel commercial strategy is heading and where operating models have yet to catch up." Lisa Murphy, HEDNA's president, was blunter about why the research exists at all: "Unbiased data beats hype."

The guest-facing version of the same problem is easier to see. Hotels have rushed to deploy AI messaging and concierge bots, and Skift Research's latest State of Travel report found that 62 per cent of global travellers are now familiar with AI trip-planning tools. Cornell's Center for Hospitality Research, surveying more than a thousand American travellers across budget, premium and luxury tiers, found that AI still ranks only fourth among the tools people actually use to plan a trip, with accuracy the single biggest reason people hesitate. A chatbot that answers confidently and wrongly, because it was trained on a static knowledge base rather than connected to the property's live booking and housekeeping systems, does more damage to trust than no chatbot at all.

This is where the RateGain research offers its most useful, and least publicised, finding. The hotels showing the clearest gains are not the giants with the biggest AI budgets. They are mid-sized chains: large enough to invest in real systems and specialist staff, small enough to avoid the tangle of legacy platforms and internal fiefdoms that slows a global brand to a crawl. Mid-sized operators reported the strongest cross-functional alignment, the most mature AI governance and the highest share of genuine, measurable reductions in manual work.

That is, more or less, a description of CityBlue's position, and of the wider East African hotel landscape it sits inside: a handful of markets, a portfolio in the dozens rather than the hundreds, ownership close enough to operations to make one decision and apply it everywhere at once. It is also why the Profitroom findings and the RateGain findings are not in tension, despite appearances. Africa did not skip the hard part of AI adoption. It simply reached the hard part first, because it moved faster than everyone else to get there. The h2c study found that 47 per cent of MEA hotel chains already report internal data silos limiting what their AI can do, the highest share of any region in the survey and well above Europe's 28 per cent. "African hotels have leapfrogged their competitors in embracing AI but now face the challenge of breaking down internal data barriers, and making sure platforms and systems are aligned," Raiter-Łuksza said. "The next frontier for African hospitality isn't just about adopting more AI tools but creating unified data strategies that deliver consistent guest experiences and measurable business outcomes."

The commercial argument for getting this right, rather than simply buying more software, is spreading beyond the technology vendors making the case for their own products. Writing in Business Daily Africa in August, Anton Gillis, co-founder and chief executive of the Kenyan hospitality group HAMAC, argued that digital transformation in the country's hotel sector had moved from an efficiency question to a competitiveness one. At the Hotel & Hospitality Expo Africa in Cape Town in June, Brett Hendricks, chief executive of Motsamayi Tourism, described the same shift from the operator's side: "Tech adoption has shifted from 'nice-to-have' to 'must-invest'," he said, citing rising labour costs, thin margins and the difficulty of finding skilled staff as the real forces pushing hotels towards automation, not novelty.

None of this argues against African hotels having moved early. It argues against mistaking the move for the destination. Wei Manfredi, IHG's senior vice-president for AI and architecture, made the point from the other end of the industry at Skift's Data and AI Summit in New York in June, when she described what her own giant, well-resourced chain had got wrong about the problem. "We are technologists," she said. "We always thought, given how powerful AI is, it's all about technology. It's not. It's really the culture and the people." The unglamorous work, she added, was "the boring stuff": clean data, working interfaces, systems that talk to each other before anyone lets an algorithm make a decision.

For an owner in Nairobi, Kigali or Zanzibar weighing an AI pitch this quarter, the question worth asking a vendor is not whether the system uses the latest model, but whether it can prove, with a number, that it has cut someone's workload by more than 30 per cent at a property that looks like theirs. On the evidence gathered so far, most vendors cannot answer that question yet, in Nairobi or in New York. The region that adopted first now has the clearest shot at being the region that proves it first, but only if it treats data integration as the actual project, and the chatbot as the visible tip of it.

Verjee's announcement in March was not proof that AI works in African hospitality. It was proof that African hospitality has stopped waiting to find out.