Telecoms Launch 'Open Telco AI' Standard
Global telecom industry body GSMA has launched a new initiative called 'Open Telco AI' to create shared, standardized AI tools for the industry. The collaboration aims to speed up innovation in network management and customer service by making next-gen telecom infrastructure both AI-native and interoperable.
The push for this standard stems from a significant performance gap; general-purpose AI models often struggle to interpret complex network data and automate telecom operations with the necessary accuracy. A GSMA Intelligence report noted that only 16% of generative AI deployments in the telecom sector have been applied to network operations, highlighting the need for specialized tools. AT&T and AMD are named as founding supporters, making substantial initial contributions. AT&T is releasing a family of open, hardware-agnostic telco models, while AMD is providing the essential GPU compute power for training and fine-tuning these AI systems through its platforms and cloud partner TensorWave. The collaboration extends far beyond the founding partners, involving a broad coalition of operators, vendors, and academic institutions. Supporters include major global players like China Mobile, Deutsche Telekom, Google Cloud, IBM, Nvidia, Orange, SK Telecom, Telefónica, and Vodafone, who will contribute data, models, and use cases. A key component of the initiative is a new online portal, GSMA.com/open-telco-ai, which will provide access to these open models, specialized datasets, and tools. This shared foundation is intended to prevent duplicated efforts and foster a collaborative environment for innovation across the industry. To measure the effectiveness and progress of these new AI models, the initiative has established a "Telco Capability Index." This benchmark will track model performance across a range of telecom-specific tasks, providing a clear metric for their development and readiness for real-world network environments.
