Telecommunications Companies Adopt Open-Source AI Models for Network Control and Customization

Telecom operators are increasingly leveraging open-source AI models to build trustworthy and customizable artificial intelligence systems for critical operations ranging from network automation to customer service. Open models provide telecommunications companies with cost advantages, deployment flexibility across multiple infrastructure environments, and the ability to fine-tune systems using proprietary network and operational data. A recent NVIDIA industry report found that 89 percent of telecom respondents view open-source models and software as strategically important to their AI initiatives.
Telecommunications companies are shifting toward open-source AI foundations to balance innovation with operational control. The approach allows operators to access cutting-edge AI capabilities while maintaining the flexibility to customize systems for their unique network environments and regulatory requirements. Organizations like SoftBank are building proprietary telecom-specific models on top of open foundations, combining public research advances with decades of accumulated domain expertise in network operations and management.
NVIDIA's development of specialized models like the Nemotron 3 Large Telco Model represents an effort to bridge open-source AI with telecom-specific requirements. By providing pre-tuned models trained on telecom datasets alongside detailed fine-tuning recipes, the vendor enables operators to accelerate deployment while retaining the ability to adapt systems to their particular infrastructure, customer bases, and operational procedures.
Widespread adoption of customizable AI systems in telecommunications infrastructure could enhance network efficiency and service quality while potentially reducing operational costs for providers. However, this trend may also concentrate decision-making power in fewer technology vendors and create dependency on specialized AI expertise. The emphasis on local customization and edge deployment could affect data sovereignty and privacy practices across different regions, while competitive pressures might accelerate automation's impact on telecom workforce composition.