Today, telcos are implementing GenAI in a range of modes; employees are using open source LLMs, they are internally training models for specific use cases and also deploying vendor models. However, the investment in GenAI remains below 20% of the AI/ML spend.

Things impeding progress include:

• Concerns of GenAI going native and being left to interact with customers unsupervised

• Cost and environmental sustainability of the potential future AI compute required

• Ongoing and potentially onerous regulations that are costly and stifle innovation

• An unanswered question as to who owns the regulatory burden in an open source LLM when it is used by a Telco

But all of this has not stopped practical experience with GenAI growing in both telcos and vendors. A 2024 Nvidia survey found that investing in customer experience optimization was
the most popular AI use case for 2023 (49% of respondents). The survey also indicated that for GenAI customer service and support were the most popular investments.