T-Mobile is putting more attention on artificial intelligence again, but this time the focus is less about buzzwords and more about how AI could be used behind the scenes to improve wireless service and customer experiences. The company, alongside Deutsche Telekom, has announced the winners of its 2026 T-Challenge, a global competition built around finding startups with AI ideas that could actually work in telecom.

T-Mobile announced that this year’s challenge centered on the push toward what the companies call an “AI-native telco,” meaning a wireless business that uses AI directly in network operations and customer-facing services instead of treating it like a side project. T-Mobile says the goal was to move beyond experimentation and highlight solutions that can improve performance, efficiency, connectivity, and the overall customer experience in practical ways.

The top prize went to Stanford University in the U.S., while Cubig of South Korea took second place and Germany-based Daisytuner landed third. A special award was also given to zTouch in the U.S. T-Mobile and Deutsche Telekom said the winning teams were chosen from hundreds of applicants and stood out for building AI tools that could be used in real telecom environments, with work touching areas such as network intelligence, automation, and better customer experiences.

For T-Mobile, the announcement is really a reminder of where the company wants to head next: using AI not just as a flashy talking point, but as part of the way its network and services operate day to day. The T-Challenge also remains part of T-Mobile’s broader collaboration with Deutsche Telekom to work with startups and speed up new ideas that could shape the future of connectivity. 

More details on the winners are available at T-Challenge’s official site, t-challenge.com.