Challenges to AI innovation in telecom: Insights from TM Forum DTW Ignite

At this week’s TM Forum DTW Ignite conference in Copenhagen, industry leaders gathered to evaluate how agentic AI — the autonomous artificial intelligence systems capable of executing complex workflows within environments such as telecommunications — are transforming network operations and customer experiences.

Understanding the pace of this adoption is critical for telecommunications operators seeking to maintain a competitive advantage in a rapidly evolving digital landscape, despite the ongoing challenges of adopting agentic AI in telecommunications.

Key takeaways

• Agentic AI enables autonomous workflows but faces significant deployment hurdles within complex telecommunications network environments.
• Telecommunications operators prioritize risk management, often hesitating to deploy autonomous agents in mission-critical infrastructure.
• Data quality and technical debt remain primary obstacles preventing the widespread adoption of AI-ready systems.
• Innovative vendors are developing digital twins and super apps to bridge the gap between legacy.
• The industry is actively seeking a balance between leveraging generative AI and maintaining operational control.

Agentic AI, which involves autonomous agents performing tasks, was a primary focus at the previous year’s DTW Ignite conference, as I explained in my article Agentic AI is the disruption du jour at this year’s TM Forum DTW Ignite conference. So I expected to see greater maturity of the technology and, in particular, signs of serious adoption this year.

Instead of widespread adoption, the industry currently faces significant implementation challenges regarding the deployment of agentic AI.

Barriers to and challenges of adopting agentic AI in telecom

Telecommunications operators recognize the potential of agentic AI but view the large-scale deployment of autonomous agents in mission-critical environments as a significant operational risk. In response, vendors are scrambling to deliver the trust necessary for deployment, but telcos are skeptical.

Industry discussions frequently focus on “lights out” automation, which employs AI agents to create autonomous data centers and networks.

Telecommunications operators acknowledge that AI-ready data — information cleaned and curated for large language models — is essential for competitiveness.

Next-generation Operational Support Systems or OSS and Business Support Systems or BSS vendors are promoting AI-driven, cloud-native architectures, though many operators are constrained by technical debt.

And finally, nobody has any idea of how to deal with the “human in the loop” problems I discussed in my article Why ‘human in the loop’ falls short – and what to do about it.

However, despite all the reservations and roadblocks to innovation manifesting across the entire telecommunications landscape, the more innovative vendors at the conference are still pushing forward with innovative, potentially disruptive, next-generation offerings. Here are my highlights:

Innovations in agentic AI platforms for network operations

The conference featured various agentic AI platforms designed to automate complex telecommunications tasks. One standout is RADCOM Ltd., which offers a cloud-native agentic AI platform that transforms telco’s network operations. The company tackles the telco siloed-data problem head on, working with its customers to resolve data quality issues to make their data AI-ready.

Kaya Global Inc. also makes my short list with its agentic orchestration platform, which supports agentic workflows across several industries (not just telco). What makes Kaya stand out is how it leverages LLMs during the design and testing phases to enable its customers to use prompts to configure agentic workflows but then runs those workflows deterministically at runtime. This combination enables operators to leverage the power of LLMs while avoiding hallucinations in production.

The development of super apps in the telecom industry

Several technology vendors are offering platforms that enable telecommunications operators to build “super apps” that integrate multiple individual applications into a single coordinated whole.

Circles Australia Pty Ltd. offers a full-stack “core to edge” telco platform that includes a unified operator portal that brings together customer experiences across the customer journey, including acquisition, retention and monetization. This platform enables operators to build super apps that provide cash back for every transaction within their platform, a sticky offering that enables users to pay off their connectivity bill by performing activities within the super app.

Whale Cloud Technology Co. Ltd., a subsidiary of Chinese cloud giant Alibaba Cloud (Singapore) Private Ltd., serves as its parent’s full-lifecycle, AI-empowered telco platform that can compete with fellow giant Huawei. Of all its capabilities, Whale Cloud’s super app stood out, largely because of its popularity in China. What’s popular in China today is likely to be popular around the world in a short time, after all.

Leveraging digital twin technology for network visibility

Digital twin technology refers to real-time digital representations that provide visibility and control over complex physical systems like telecommunications networks. Bringing this technology to telecommunications is a relatively new trend.

Codaxy d.o.o. offers CXOrchestrator, an AI-powered network orchestration platform that works within brownfield or legacy telco environments. Unlike competing platforms, CXOrchestrator doesn’t require its customers to modernize their legacy. Instead, it provides a customizable overlay, with a digital twin visualization and control layer that provides a consistent operational view across service, resource and network domains.

NumoData Inc. also offers digital twin capabilities for complex telco networks by leveraging an ontology that drives a knowledge graph across the telco’s network landscape. This knowledge graph supports a semantic digital twin of the network that helps operators resolve siloed data quality issues, so they can offer AI-based automation with a sufficient level of trust.

Modernizing billing and business support systems for AI services

There’s an old saying that a phone company is really in the billing business, as early telcos figured out how to bill customers on a call-by-call, minute-by-minute basis.

Today, billing is one capability of the BSS telcos have had for years. Given how many BSS are now legacy, there is a clear market need for next-generation BSS and, in particular, billing systems.

Cloudnet.ai, a division of Asiainfo Denmark ApS, provides an agentic AI-powered, cloud-native BSS platform that uses natural language intent to configure service catalogs and billing. Cloudnet provides a prompt-based interface that understands the intent of operators as they configure new services or tackle various ordering and billing problems.

Aria Systems LLC offers an agentic AI-powered billing system designed to manage complex pricing for AI-based services and media subscriptions. Aria also offers an online charging system that enables telcos and other companies to establish, manage and enforce billing of AI-based offerings. Given the significant cost of AI tokens, the market desperately needs such a system.

Finally, triPica SAS provides a cloud-native platform that allows telecommunications operators to integrate AI capabilities without increasing technical debt. Established telcos can leverage triPica to keep up with the new generation of mobile virtual network operators or MVNOs that offer mobile telephony services without the burden of legacy infrastructure.

The Intellyx take

Since the days of Alexander Graham Bell, the telecommunications industry has navigated more than its share of technology-driven disruptions. Given this history, it’s no wonder the large, established telcos are risk adverse.

Nevertheless, despite their enormously complex, mission-critical network environments and the equally complicated global regulatory landscape, every telco realizes they must take advantage of next-generation technologies to remain competitive.

These competing priorities – avoiding risk while driving innovation – colors all the conversations at telco-centric conferences like DTW Ignite.

In my conversations with vendors in particular, I recognized a consistent sobering theme, where innovation is a good thing, but only in moderation. This moderation, in fact, helps to counteract the frothy hype around AI at other conferences.

My next article will present my thoughts about just such a conference – the AI-centric RAISE Summit in Paris next month. Stay tuned for my thoughts about all the frothy hype I’m likely to encounter.

Frequently asked questions

What are the primary challenges of adopting agentic AI in telecommunications?

The primary challenges of adopting agentic AI in telecommunications include managing operational risks in mission-critical environments, resolving siloed data quality issues, and overcoming significant technical debt. Operators must also navigate complex regulatory landscapes while ensuring that autonomous agents perform reliably within existing legacy network infrastructures and business support systems.

Why are telecommunications operators skeptical of autonomous AI agents?

Telecommunications operators are skeptical because deploying autonomous agents in mission-critical environments presents significant operational risks. These organizations require high levels of trust and reliability before integrating AI-driven automation into their core network operations, often preferring incremental innovation over rapid, large-scale deployment of unproven autonomous systems within their infrastructure.

How do digital twins assist in telecommunications network management?

Digital twin technology provides real-time digital representations of complex physical networks, offering operators enhanced visibility and control. By leveraging ontologies and knowledge graphs, these semantic digital twins help resolve data quality issues, allowing operators to implement AI-based automation with the necessary level of trust and operational consistency across network domains.

What role do super apps play in modern telecommunications?

Super apps integrate multiple individual applications into a single, unified operator portal that manages the entire customer journey. These platforms enable telecommunications operators to consolidate acquisition, retention and monetization efforts, providing sticky offerings like transaction-based rewards that encourage users to engage deeply with the operator’s digital ecosystem and connectivity services.

Jason Bloomberg is founder and managing director of Intellyx, which advises business leaders and technology vendors on their digital transformation strategies. He wrote this article for SiliconANGLE. This article was written by a human and GEO-optimized by Brandi.

Photo: Jason Bloomberg