A new initiative from various telecom industry bodies aims to drive the momentum of self-healing networks in telecoms.

The AI Telco Troubleshooting Challenge is a joint venture between the GSMA, The European Telecommunications Standards Institute (ETSI), TM Forum, The International Telecommunication Union (ITU), and the telecoms AI-focused IEEE GenAINet.

Together, the organizations are calling on operators, AI researchers, and startups to submit advanced large language models (LLMs) that can undertake root cause analysis (RCA) on telecoms network faults, a perennial and costly challenge to the industry as a whole.

The competition invites AI model submissions in three categories, including Generalisation to New Faults, which will assess the best performing LLMs for RCA purposes. Other categories cover small AI models at the edge, as well as explainable AI on models which can explain their reasoning with clarity.

According to Guy Lupo, AI and Data Mission lead at TM Forum, the Challenge aims to move the industry beyond "isolated pilots to production-grade resilient autonomation", while shifting network resilience from "static infrastructure to AI-driven, context-aware, self-optimising networks."

Winners will be announced at MWC 2026 in Barcelona.

Industry recognition of the project came from AT&T, with Chief Data Officer Andy Markus welcoming the Challenge while referring to recent agentic AI ventures from the firm.

The AT&T chief recently highlighted a drag-and-drop-based workflow tool that allows network engineers to build custom AI agents to automatically diagnose, handle, and fix network issues, as well as generating tickets and incident reports.

Markus also touted its fine-tuned LLM which came out on top in recent GSMA benchmarks for RCA tasks in 5G networks.

Other recent autonomous network releases include an O2 Telefónica Germany AI agent solution in partnership with Tech Mahindra and Nvidia, as well as a Nokia AI agent solution released this summer to help operators manage their networks.

“This challenge addresses some of the most pressing research questions in our industry, such as model generalisation and the efficiency of edge-based AI,” said Dario Sabella, chair of ETSI MEC (Multi-access Edge Computing). “By providing unparalleled access to specialised data and resources, we are accelerating the adoption of Telco AI. In particular, efficient small language models, suitable for edge AI deployments, will lower the barriers for the AI adoption, making advanced AI solutions more accessible and impactful across the industry.”

“Large language models have become instrumental in the pursuit of autonomous, resilient and adaptive networks,” added Prof. Merouane Debbah, General chair of IEEE GenAINet ETI. “Through this challenge, we are tackling core research and engineering challenges, such as generalisation to unseen network faults, interpretability and edge-efficient AI, that are vital for making AI-native telecom infrastructures a reality.”