Industry group TM Forum joined forces with Amazon Web Services (AWS) to develop a generative AI (genAI) tool to help operators benchmark, prioritise and deploy the technology.

The Generative AI Maturity Interactive Tool (GAMIT) uses anonymised, aggregated data from more than 240 AI-decision makers at communication service providers (CSPs) worldwide.

It enables operators to benchmark their genAI maturity against regional peers and global leaders to help identify priority use cases, along with steering the development of use cases to production at scale.

Initial results from the data used to build GAMIT show genAI adoption by operators remains in the early stages.

TM Forum stated operators are struggling to move beyond bolt-on AI applications to integrating it as a native element across their operations. The most common barrier is a lack of accuracy in proof-of-concepts for generative iterations of the technology.

Early TM Forum data indicates 25 per cent of CSPs feel equipped to use advanced techniques such as fine-tuning, retrieval-augmented generation (RAG) and prompt engineering.

It also found 33 per cent of CSPs have a CXO in place dedicated to building a strategy around the technology and 14 per cent have more than ten genAI use cases in production.

AWS and TM Forum stated GAMIT is designed to overcome those issues to help operators realise the full potential of the technology.

GAMIT covers technology understanding and maturity; organisation, responsibilities and skills; data readiness and availability; governance, privacy, compliance and security; business objectives; and putting AI uses cases into production.

TM Forum CTO George Glass stated GAMIT addresses the challenges operators face by giving them “a clear benchmark of their generative AI maturity and a practical roadmap to scale it across their entire organisation”.

In February, consultancy McKinsey and Co estimated genAI could deliver close to $100 million in incremental value to companies “at the forefront” of genAI, and between $140 million and $180 million in productivity gains “above what could be unlocked by traditional AI”.