Major advancements in artificial intelligence are increasing telco stakeholder expectations. This report assesses where telcos are on their AI adoption journeys and provides recommendations for how they can accelerate their progress.

A deep dive into the progress towards an AI-driven future

The use of AI in telecoms is not a new development. Telcos have been leveraging the technology across a variety of areas for several years. For example, AI has already been used to support intent-based operations, chatbots and customer churn analysis. For some time, telcos have also been working with vendor partners to integrate ML (machine learning) and predictive AI into their solutions to improve offerings and customer experience.

However, there has been a surge in interest on the topic due to the release of ChatGPT and the subsequent obsession with generative AI. Couple this with recent strides in AI design and infrastructure that have granted telcos the opportunity to access and generate value from a plethora of previously inaccessible data and it is evident why telcos are focused on AI as a means of reinvigorating profitability growth. This focus is being compounded by increasing pressure from investors and other telco stakeholders for them to make and communicate their progress with their AI journeys.

This focus seems to be warranted, and a recent STL report suggested that “an average telco can save the equivalent of 5% of its revenue (USD816 million) annually by deploying automation, analytics and AI in its network and OSS processes.” Separate estimates predict that generative AI could benefit the whole industry by USD60–100 billion annually. Simply considering these two independent estimates offers a compelling rationale for telcos to pursue AI.

While the potential benefits are clear, telcos face several barriers to success with AI that other technology companies are not faced with to the same extent:

• Telcos are more financially resource constrained than techcos and therefore have a lower risk tolerance when making new investments, particularly when existing network expansion obligations are considered.
• Telcos typically have legacy, siloed systems and infrastructure that makes it harder to capitalise on AI tools, which need a high volume and quality of data to produce useful output.
• Telcos struggle to access the skills and capabilities required to build and leverage AI solutions, as talent is often attracted to hyperscalers and app/platform developers, as evidenced by STL’s Future Skills Tracker. This can create dependencies on hyperscalers, particularly when telcos need large foundation models.
• Telcos often operate with heightened regulatory scrutiny and higher privacy expectations when it comes to data which is the lifeblood of AI models.

If you are not a subscriber, enter your details below to download an extract of the summary deck

As the chart below illustrates, our survey identified a lack of consensus over which challenges are most significant to telcos. It shows that the range in average score of each category of challenge from our 104 survey responses is narrow around the median score of three. Respondents were also surveyed on challenges within each category and the narrow range remained true for four out of the five challenges. Interestingly, there was not a significant variation in the rankings between respondents with technical versus commercial roles. This suggests that telcos are struggling to prioritise which challenges are most significant and therefore which they should tackle first.

The lack of consensus over which challenges are most significant highlights the scale of the task for telcos implementing AI

These challenges and the difficulties associated with prioritising them is leading telco decision makers to hesitate. To help executives address the challenges they face with AI and overcome the (potential) paralysis they face, this report seeks to answer these questions:

1. What strategic direction should telcos take to maximise the positive impact of AI adoption?
2. What capabilities do telcos need to execute their AI strategy and overcome challenges?

Table of contents

• Executive Summary
• Background on our research
• Introduction
• Adopt customer-centric AI strategies
• Embody agility to accelerate use case maturity
• Build a cross-domain centre of excellence (CoE) to accelerate AI adoption
• Take an open approach to plug capability gaps
• Learn from the success of others
• Conclusion

Related research

• Telco generative AI adoption tracker
• Organisational foundations for responsible AI
• DTW24: Generative AI brings both hype and hope
• Finding value from AI, analytics and automation (A3) in the telco network

Author

Will Davies

Consultant

Will Davies is a consultant at STL Partners specialising in Sustainability.