Last week’s TM Forum Innovate Americas trade show in Dallas, Texas brought together telcos and their partners from across the Americas to discuss how AI and automation can power the future of telecommunications. As one might expect at an event focused on AI, much of the discussion centred around Generative AI (Gen AI), specifically how telcos can leverage this emerging capability. Numerous presentations were given by AI vendors and telcos, highlighting experimental use cases that have been implemented—typically as standalone pilots or proofs of concept (PoCs). Tangible results remain limited at this stage, which is understandable given the early stages of learning how to effectively use this fast-emerging technology and avoid common pitfalls.
Generative AI in Telecom: Early Use Cases and Opportunities
The Gen AI case studies presented at the show were diverse, but the most promising opportunities for telecom operators seem to lie in the customer service domain. Pilots have been conducted using AI agents to handle frontline enquiries, and early results are positive. The quality of responses to customer queries appears much improved compared to previous business-rule or machine-learning-based chatbots. Integration with natural voice AI services, such as voice.ai, has also significantly upgraded the user experience. Over time, AI agents will likely handle more complex queries, such as Level 2 and Level 3 service issues, but for now, they are primarily being used to manage a growing portion of frontline customer service tasks.
Other use cases worth noting include AI agents managing back-office workloads. However, caution is still required in areas where 100% accuracy is essential. Since these agents ultimately rely on predictive algorithms, systems of record remain necessary to maintain a “single version of the truth” in critical operations.
Gen AI in Marketing: Personalisation and Dynamic Content Creation
In the marketing domain, where Ikue operates, there were some intriguing early-stage experiments. Gen AI is being used for next best offer decisioning and experience personalisation. While we are still in the early stages of this development, the use of Gen AI to dynamically render digital creative content at the point of customer interaction is gaining traction and seems like a natural early use case for marketers to explore.
In summary, there were many fascinating discussions at the show, with much to learn for those interested in AI’s potential in telecom.
Key Takeaways:
- Many proof-of-concept (PoC) projects are underway across numerous telcos, but few have reached genuine scale. This is largely due to the novelty of the technology, but if we revisit this in 12 months, the landscape is likely to have changed significantly.
- The coordination of pilots across different business units within telcos is somewhat fragmented. While the desire to learn quickly is commendable—especially given the historically slow pace of large telcos—the lack of governance across Gen AI programmes might lead to missed organisational learning opportunities. There is also a question of control when it comes to managing these programmes—but more on that below.
- Access to the right data is crucial for enabling large language models (LLMs) to learn from proprietary data. However, assembling that data into a single location and formatting it correctly for LLMs to process at low latency is a challenge many telcos are struggling with. This is an area where we at Ikue believe we can provide valuable assistance to help telcos accelerate their AI implementations.
- A larger issue, not just limited to telco, is the question of data permissions. How can telcos safely feed their data into a public LLM while safeguarding customer privacy and intellectual property (IP)? Do they need to train their own models instead? If so, this approach could become prohibitively expensive, raising concerns about the long-term business case for such investments.
Concluding Thoughts: What’s Next for Telcos and Gen AI?
Stepping back to conclude, the timing of this show, with its focus on AI and Automation, was perfect, given the immense attention that Gen AI is receiving across multiple industries. Hosting the event in the U.S. was also strategic, as many of the current leaders in the AI space, such as OpenAI, Anthropic, and Meta, are based there.
It will be interesting to see if TM Forum takes this concept to other geographies and if telcos in different regions are pursuing the same Gen AI applications or experimenting with different use cases.
For those in the telecommunications industry, Innovate Americas provided an excellent primer on the emerging possibilities of AI and automation.
Want to learn more about Ikue? Contact us!