Voyage of discovery

Technology, especially AI, is reshaping how the travel industry operates – from bookings to customer experience

Voyage of discovery

The most important changes in travel technology may be the ones travellers never become aware of. Behind a seemingly simple request for a hotel or flight, a growing layer of artificial intelligence-enabled infrastructure is sorting, ranking and connecting information from multiple systems.

Travelport, a global travel technology company facilitating bookings for millions of travellers worldwide, is rebuilding that infrastructure around AI with TripServices, a cloud-native platform designed to connect flights and stays through a single API (application programming interface), with AI-powered content curation, search and ranking built into the system. Its partnerships with AI developer Anthropic and multinational IT consulting company Cognizant are intended to bring AI further into the workflows connecting travellers with travel suppliers.

At the consumer end, Expedia – a US travel tech company that operates travel fare aggregators and travel metasearch engines – is betting that AI-powered, conversational trip planning will eventually replace conventional searches as the starting point for booking a holiday. In July, the company acquired Layla, an AI-native trip-planning platform that allows travellers to discover destinations, build personalised itineraries and plan trips through dialogue, including flights, accommodation, activities, dining and transport.

Similarly, Accor Group – the French multi-national hospitality giant behind global hotel brands such as Fairmont, Novotel and Ibis – has rolled out ALL (Accor Live Limitless, the title of its loyalty programme) Concierge, a conversational AI travel companion that can recommend hotels, retrieve real-time property information, make and manage reservations and provide post-stay support. The company says it learnt from more than 1 million guest conversations during the pilot.

In SA, technology is being used to tackle basic barriers that can prevent a tourism transaction from happening at all – particularly the difficulty and cost of making international payments. SA fintech TurnStay, which builds payment infrastructure for travel businesses, says it has processed more than R1 billion in 2026. Its platform is designed for tour operators, safari lodges and villa-rental agencies, offering multi-currency payments, payment links, virtual cards and other tools aimed at reducing payment costs and improving the likelihood that transactions succeed. There are signs that this kind of infrastructure can make a measurable difference. Michael Aldrin, finance manager at private nature reserve Thornybush, says TurnStay has reduced payment failures, expanded the currencies available to guests and lowered fees.

‘What sets South Africa apart is that our digital foundations often need attention before AI can deliver its full value,’ says Leigh Myles, regional director for Africa at cloud-based booking and marketing platform Profitroom. ‘Payments are a good example. The average payment success rate on South African hotel booking engines is around 50%, compared with 70 to 90% globally. That has improved from roughly 35% in 2022, so progress is happening, but it shows how much friction still exists in the basic online journey. AI layered on top of a slow website, a clunky booking engine or an unreliable checkout won’t fix the problem. So the opportunity here is twofold: get the fundamentals right, and use AI to leapfrog where we can.’

Adoption of AI and tech solutions across the national industry is ‘enthusiastic but uneven’, says Myles. ‘There’s real curiosity across the industry, and some properties are experimenting with AI for content, guest communication and revenue decisions. But for most South African tourism businesses, AI is still something they’re talking about rather than something built into how they operate.’

That’s not unique to SA, she adds. ‘Globally, the industry is [lagging] behind its customers.’

This is supported by recent market research from MMGY Travel Intelligence, which found a growing disconnect between traveller behaviour and travel industry readiness when it comes to AI. The study found that 54% of travellers have already used AI to plan travel and 75% intend to use AI for future trips. Meanwhile, only 22% of destination marketing organisations (DMOs) currently offer AI trip-planning capabilities on their websites, and just 26% have a documented AI strategy. With 72% of DMO leaders saying organisations that fail to adopt AI within the next two years risk becoming irrelevant in destination discovery, the industry is at a crossroads with a gap that is becoming increasingly difficult to ignore.

For AI ethics, governance and policy specialist Nazareen Ebrahim, AI adoption in SA tourism is best understood across two dimensions: private-sector adoption and public-sector regulatory and industry support. ‘At a private-sector level, South Africa appears broadly aligned with the direction of global adoption. There is a clear appetite for AI, particularly in areas such as customer service, marketing and content creation, data analysis, personalisation and operational efficiency. As elsewhere globally, however, organisations are at very different levels of maturity: some are integrating AI into workflows, while others are still experimenting with tools, building internal capability and identifying appropriate use cases.’

This reflects a broader global pattern. A 2025 global study from hospitality consultancy H2C on AI and automation in hospitality, covering 171 hotel chains and more than 11 000 properties worldwide, found that 78% of hotel chains were already using AI, and 89% planned to expand their use within the following 12–24 months, yet only 6% had a comprehensive company-wide AI strategy.

This is an important distinction, according to Ebrahim: widespread use of AI tools should not automatically be equated with organisational AI maturity.

‘Where South Africa has more ground to cover is in the broader enabling environment. From a public-sector, regulatory and industry-support perspective, there is still a need for stronger sector-specific guidance, governance frameworks, AI literacy and skills development, and practical support that enables tourism businesses – particularly smaller operators – to adopt AI responsibly. I would not characterise South African tourism as simply “behind” globally. At the level of appetite and experimentation, we are participating in the same transition taking place internationally. The greater challenge is ensuring that adoption evolves from individual tool use into strategic, responsible and sustainable AI adoption supported by an enabling industry ecosystem.’

There are signs that some of that wider support is beginning to take shape.

In September 2025, SA’s inaugural G20 Tourism Hackathon brought together 48 students from 21 higher-education institutions to develop technology-based solutions for the sector. The winning concepts included a holographic platform designed to help rural communities showcase their heritage, an AI-supported map intended to make rural tourism businesses more visible to travellers, and a platform combining virtual reality, AI and Web3 (a decentralised internet built around technologies such as blockchain) that allows international tourists to explore rural villages virtually. The projects were intended to move beyond ideas towards commercialisation, with funding from the Department of Tourism and its partners.

In Cape Town, the response has been more industry-focused. Cape Town Tourism established its AI in Tourism Hub in August 2025, bringing together tourism businesses, technology specialists, academics and AI policy bodies to develop practical applications for the sector. Its areas of focus include AI for marketing, visitor safety, sustainability and operational efficiency.

But as adoption grows, the question becomes less about whether tourism businesses should be using AI and more about where it is genuinely useful – and how they can distinguish measurable value from the hype surrounding the technology.

‘Start with the business problem,’ says David Frost, CEO of the Southern Africa Tourism Services Association. ‘Where are you losing time, where are customers dropping out, and where would a faster response or a better decision make a difference? If AI improves turnaround times, cuts repetitive admin, helps staff understand a customer better or lifts conversion, you can measure whether it is working. If you cannot measure it, be sceptical.’

The danger, he says, lies in automating the parts of tourism that customers are paying you for. ‘In inbound travel, that is often local knowledge, judgement and the confidence that someone can deal with a complex trip when things change.’

He recommends keeping a few basic guardrails, such as checking the accuracy of everything AI produces and keeping a human involved wherever getting it wrong would be costly. ‘An itinerary with the wrong transfer time quickly turns a time saving into a complaint.’

Looking further ahead, AI is likely to become less of a standalone tool and more of an underlying layer across the travel journey – changing how people discover destinations, compare options and make bookings.

‘The biggest change will be in how travellers find and buy travel,’ says Frost. ‘Today people still move between search engines, destination websites, booking platforms, tour operators and suppliers. Over the next five to 10 years, AI is likely to sit across much more of that journey, narrowing choices, planning trips and making bookings.’

This makes discoverability a commercial issue. A good product is not enough if AI tools cannot find it, understand it and recommend it, he says. ‘Businesses need accurate, current information on their own websites and on every booking platform and review site these tools draw from. Many still need to get the basics of search right before they think about AI.’

For Ebrahim, the opportunity for SA tourism is not only to become more technologically advanced; it is also to ensure that the digital representation of SA’s tourism assets is developed in a way that is ethical, locally beneficial and respectful of the people and communities from whom much of that knowledge originates.

‘That, to me, is where the real conversation over the next decade sits – not simply what AI can do for tourism, but how we ensure South Africa retains agency over the value created from its own tourism data and knowledge.’

By Robyn Maclarty
Image: iStock