Grab Launches AI Assistant to Make Booking Activities Easier in One Conversation

Philipp Kandal, Chief Product Officer of Grab, at the GrabX 2026 event (photo: Rana/VOI)

JAKARTA - Grab has introduced a new Artificial Intelligence (AI) innovation on its platform. Now, the travel application is integrated with a simple chatbot named Grab AI Assistant.

As Philipp Kandal, Chief Product Officer of Grab, said at the GrabX 2026 event, this new innovation is a digital assistant designed to make user activities easier. This chatbot can simplify user needs more practically.

Grab AI Assistant utilizes various AI-based models, including OpenAI and Anthropic. These AI models are claimed to have been equipped with an understanding of user habits and preferences in Southeast Asia.

"We are working with OpenAI, Anthropic, and many other companies and open-source models that we are looking for for various countries," Philipp told the media on Wednesday, March 8.

This digital assistant is able to plan various activities, ranging from team meals to restaurant reservations, in one conversation. Users only need to provide details of the number of members and dietary restrictions, then the system will provide recommendations and reservations automatically.

Philipp explained that this technology was built in close collaboration with major AI companies around the world. Grab also stated that they use the best models that have been customized to suit the needs of users in each country.

This AI chatbot has been tested in various languages, including English and Indonesian. Testing these languages is done to ensure the accuracy of the service at the local level.

"As an AI assistant, a smart companion, he will tell users, and give advice (about) where they should move next," continued Philip. He also believes that this Ai integration will also help partners.

This latest AI technology is expected to increase the revenue of paramitra through a more efficient and targeted system. During collaboration with AI companies, Grab also helps them in providing data.

"So, what we always do is we give feedback to partners and data to help and prove their models for everyone," added Philip.