OpenAI has acquired Odessia: is ChatGPT preparing a new way to plan and book travel?
The acquisition of the AI travel startup Odessia raises the question of ChatGPT's future role in travel distribution. OpenAI could develop a system that would not only recommend trips to users, but also help them with the actual organisation and booking.
On 10 November 2026, OpenAI acquired Odessia, an American startup specialising in AI-assisted travel planning and booking. The financial terms of the transaction were not disclosed, and the company has not yet presented concrete plans for integrating Odessia's technology into ChatGPT.
Odessia was founded in early 2026 by Francis Davidson, also known as the co-founder of Sonder, which operated in the short-term accommodation sector. After the acquisition, Davidson moved to OpenAI, where he will be working on developing travel-related solutions.
Although OpenAI has not announced a new tourism product, its acquisition of a specialised company demonstrates an interest in an area where artificial intelligence is already changing the way travellers find information and make decisions.
What exactly did OpenAI actually buy?
Odessia was developing an AI travel agent that enabled users to organise trips based on their own preferences, budget and other criteria.
Unlike traditional online search engines and booking platforms, the concept was based on a conversation with artificial intelligence. The user would describe the kind of trip they wanted, and the system would help choose the destination, accommodation, transport and activities.
In June 2026, the company raised six million dollars in investment capital in a funding round led by Sequoia Capital.
The standalone Odessia service has ceased operations following the takeover. Support is still available to customers with existing bookings and travel plans.
The platform's shutdown is an interesting detail, as it raises the possibility that OpenAI is not primarily interested in continuing Odessia's operations as a separate brand. One plausible explanation is that it is seeking to acquire the development expertise, technology and specialists who have already been working on the practical application of AI agents in tourism.
Such an acquisition could accelerate the development of its own tourism functionalities within existing OpenAI products.
Why is OpenAI interested in travel?
Tourism is a particularly suitable field for the development of advanced AI agents, as organising travel often involves a series of interconnected decisions.
The traveller must compare flights, choose accommodation, check booking conditions, organise transfers and align activities with the available time and budget.
Nowadays, these activities often take place across several different platforms. One is used for searching flights, another for hotels, a third for activities, while the user themselves links all the information together.
An AI agent could consolidate a significant part of that process.
For example, a user could request a seven-day trip to Japan with a specific budget, preferred hotels and several planned activities. Instead of searching various websites individually, the system could prepare a tailored travel proposal, check availability and, with the necessary approvals, assist with bookings.
The shift from recommendations to concrete actions is precisely one of the areas where Odessia's experience could be useful to OpenAI.
Could ChatGPT become a platform for booking travel?
OpenAI is already collaborating with travel companies such as Booking.com, Expedia and KAYAK. However, the ability to search for travel deals is not the same as fully managing the booking process.
For a genuine booking, it is necessary to connect to suppliers' systems, access up-to-date prices and availability, and process payments, cancellations and amendments.
These are areas where traditional tourist distributors have a well-developed infrastructure and many years of experience.
The takeover of Odessia can therefore be viewed through several possible developmental trajectories.
The first is more advanced trip planning within ChatGPT, with better integration of various tourist services.
The second is the development of AI agents that could carry out specific booking tasks on behalf of users, with clearly defined authorisations and payment controls.
The third option concerns expanding the partner ecosystem, in which ChatGPT would become the user interface, while travel companies would continue to provide product availability, booking infrastructure and support.
None of those scenarios has yet been officially confirmed as a result of the acquisition.
What could this mean for travel agencies and OTA platforms?
For the tourism industry, a particularly interesting question is who will control the first contact with the traveller in the future.
The traditional model of online distribution is based on the user visiting a specific platform, browsing the available offerings and completing the purchase there.
If AI assistants become the primary place where travellers begin planning, part of the selection process could move directly into a conversation with artificial intelligence.
This could affect several segments of the tourism business.
For OTA platforms, the importance of linking their own reservation systems with AI assistants is growing. Hotels and airlines could gain additional distribution channels, but also face new challenges in managing inventory visibility and customer relations.
Travel agencies could use advanced AI agents to prepare more complex itineraries, compare offers and automate routine tasks. However, professional support, change management and accountability for the realisation of trips remain important parts of agency operations.
It will be particularly interesting to monitor the relationship between AI platforms and existing distribution systems, including GDSs, NDC connections and direct API integrations with suppliers.
For the commercial deployment of AI agents, merely understanding the user's request will not be sufficient. Systems will need to reliably process actual tariffs, booking conditions and post-purchase changes.
The next challenge is not recommendation, but the realisation of the journey.
The development of generative artificial intelligence to date has shown how quickly AI can prepare destination suggestions, itineraries and comparisons of tourist products.
However, the real value of advanced travel agents will depend on whether they can turn such recommendations into reliable bookings.
Odessia's experience in developing AI agents and Davidson's knowledge of the tourism business could help OpenAI precisely in that area.
As of now, there is no confirmation that ChatGPT will get its own complete booking system, nor is it known when any new features might become available.
However, the acquisition suggests that OpenAI is exploring possibilities that go beyond traditional passenger information. For travel companies, this is an additional reason to monitor the development of AI distribution, particularly the ways in which their offering will be available to digital assistants and how bookings will be handled in such a model.
