Telecom Operators Adopt Open Models for AI Strategies

89% of telecom operators report the importance of open source models in their AI strategies. SoftBank Corp. is utilizing open models to enhance its AI development, as stated by Rajeev Koodli, Principal Fellow and Senior Vice President of SB Telecom America, who noted that these models enable the company to leverage its extensive network knowledge and operational expertise.

AT&T is also implementing open models in its AI strategy, with Chief Data and AI Officer Andy Markus emphasizing the need for a flexible approach that intelligently matches workloads to the right combination of performance, cost, and control.

Indosat Ooredoo Hutchison, one of Indonesia's largest telecom operators, is similarly focused on adapting open models to meet local needs. Chirag Sukhadia, the Chief Data and AI Officer, highlighted that for countries like Indonesia, the value of open models extends beyond access to powerful AI; it involves tailoring intelligence to fit local languages, cultures, and data.

NVIDIA supports these initiatives with its Nemotron family of open models, including the Nemotron 3 Large Telco Model (LTM), which boasts 30 billion parameters. AdaptKey is a partner fine-tuning NVIDIA models specifically for telecom datasets. Additionally, NVIDIA's AI Enterprise software powers its end-to-end platform for telecom AI, while the NVIDIA Agent Toolkit facilitates agent orchestration in AI workflows. This follows NVIDIA's recent release of the full recipe for fine-tuning open models, further enabling telecom operators to optimize their AI capabilities.