OTel 2.0: Dell, AT&T, AMD bring open telecom AI to production

The release of OTel 2.0 marks a significant step in the Open Telco AI initiative, which was launched earlier this year with GSMA's help to create a shared ecosystem of telecom-specific data, standards, and benchmarks. AT&T joined as a founding contributor, bringing real operator insight based on genuine network conditions. The collaboration with Dell and AMD addresses the immense computing demands of training at this scale, with AMD providing the high-performance hardware foundation and Dell integrating it with AI-optimized servers for telecom environments.

A key aspect of this deployment is that it is on-premises, keeping data under operator control throughout training with no public cloud dependency. This approach addresses governance concerns in a highly regulated industry. As training continues in the coming weeks, updated model weights will be released, and the partnership aims to close the gap between what Open Telco AI can do and what it delivers in live production environments, where reliability is non-negotiable. The open nature of the models gives operators flexibility to deploy across clouds or on-premises without vendor lock-in.