TM Forum research found 94 per cent of operators believe generative AI (GenAI) would have a significant impact on their businesses over the next one-to-five years, but cited implementation as a key challenge.

Half of the 104 senior-level respondents from 73 operators plan to create an entirely new strategy to tap GenAI. The other half that currently use AI and ML for decision-making and predictive tasks plan to add the capabilities to their existing strategies.

For the technology stack, 59 per cent plan to use an open-source cloud model and off-the-shelf large language models (LLMs) to rapidly deploy and scale GenAI.

To mitigate security concerns, 32 per cent stated they would adopt on premise models.

Privacy and security was listed by 80 per cent of the respondents as the biggest concern for using GenAI, while 74 per cent flagged a “lack of truth function” as a challenge.

The survey found 74 per cent were also worried about the legal implications from the use of public data to develop the LLMs.

Mark Newman, chief analyst for TM Forum, stated deciding between using off-the-shelf LLMs versus developing proprietary models “will largely depend on which use cases operators want to address and their appetite for investment”.

Among the survey respondents, 55 per cent indicated they had made progress in adopting GenAI, while 31 per cent are running experimental trials to develop proof-of-concepts before wide-scale deployments.

TM Forum identified customer operations as the low-hanging fruit for operators, with 92 per cent using “GenAI’s abilities to process unstructured data alongside structured data to improve customer chatbot experiences”.