Trends, insights, and a vision for scaling AI at the speed of trust
During the last week of June, Copenhagen’s Bella Center becomes the global meeting point for the telecom industry, as DTW Ignite 2026 brings together operators, vendors, standards bodies, and technologists under one roof. DTW remains a premier forum for the industry to take stock of where it is, debate where it’s going, and, critically, how to get there.
Here’s what we expect to take center stage.
Autonomous Networks: The Race to Level 4
Autonomous networks will once again be a dominant theme, and this year the conversation has materially shifted. It’s no longer hypothetical.
TM Forum’s six-level autonomy framework has been the industry’s north star since 2019, and progress toward Level 4, where near-full automation is possible across many scenarios, is becoming a reality for a small but growing number of operators. Rakuten Mobile, a RADCOM customer, is one example that has already demonstrated Level 4 at scale in a live RAN network, achieving 20% RAN energy savings through AI-driven closed-loop control. And TM Forum has noted a “significant change” in the pace of Level 4 validations among select operators heading into 2026.
The broader picture, however, is slower. TM Forum found only 21% of operators are operating at Level 3 autonomy or above overall, up only slightly from 19% the year before. While industry progress appears modest on average, the reality is more polarized: a small group of leaders is accelerating toward Level 4 autonomy and beyond, while many operators remain stuck in the earlier stages of the journey.
We expect sessions on how cross-domain autonomy can unlock the next level of scale, including how distributed agents coordinate from the RAN to the core as well as the interoperability challenges the multi-agent approach introduces. Or see RADCOM’s Catalyst “Agent Fabric: A2A-T Runtime – Phase III,” which tackles this problem by moving the industry from isolated agents exchanging loosely structured prompts to agents that negotiate tasks, share outcomes, delegate workflows, and operate with a shared operational context.
Scaling AI: From Quick Wins to Real Returns
Most operators already have some AI capabilities in production, but not all are the kind that moves the needle. The hard part is building AI that runs reliably and at telco scale, and that’s where most operators are stuck.
The disconnect between AI ambition and operational readiness is reflected clearly in the numbers.
A report in Mobile World Live revealed that 88% of telecom operators report expecting better returns from AI initiatives, yet only 12% are building what could be called engineered AI products. Most investment remains concentrated in embedded tools such as task automation, call summarization, and sentiment tagging. These are real wins, but they are not transformative. Meanwhile, the majority are still operating on legacy infrastructure not designed for the intensity of telco-grade AI, and only about a third say their infrastructure can scale with evolving business demands.
Expect many DTW discussions to shift toward the practical realities of scaling AI, starting with the data foundations required to make it work. Dedicated spotlight sessions will focus on how service providers are working to deliver real-time data so AI-driven network automation, based on robust, at-scale data platforms, can integrate seamlessly across agentic AI environments.
Trustworthy AI: From Conversation to Reality
Trust has emerged as the operating condition without which everything else stalls. Move faster than trust allows, and unforeseen risks proliferate. Move slower, and competitors pull ahead. It’s a real tension, and in 2026, it’s no longer just philosophical.
Trustworthy AI means systems that are explainable, fair, transparent, robust, and secure, and operators say they aren’t there yet. A recent Heavy Reading-RADCOM survey of telecom operators confirms it: after integration complexity, the single biggest obstacle to leveraging AI in the network is accuracy and trust in the models themselves. As AI researcher, Qiang Liu, from the University of Nebraska-Lincoln School of Computing, recently put it: “Everyone’s using AI, and they want to use it, and the performance looks great in the testbed, but they still don’t trust it.”
This is particularly acute in telecom network operations. Traditional NOC automation is mostly rule-based and open-loop, with scripts that trigger on fixed conditions but struggle with noisy signals, cross-domain dependencies, and constantly changing network behavior. As AI takes on more autonomous decision-making in live networks, the ability to explain those decisions, audit them, and maintain human oversight becomes non-negotiable.
Telco data sovereignty is part of this trust equation: operators cannot build resilient networks if they don’t trust the data or know where sensitive data flows, who can access it, or how it is governed. Expect Copenhagen to host pointed conversations about what trustworthy AI actually looks like in production, and how operators can move at the speed of trust without falling behind.
API Monetization and the TechCo Transition
Underpinning all of this is a strategic question that the entire industry is wrestling with: how does a telco become something more?
Despite historic scale and global infrastructure investments, telcos continue to experience lagging financial returns compared to the broader tech industry. The argument gaining traction is that the answer lies in the transition from a traditional connectivity provider or telco, to an AI-first technology company, coined TechCo. And network APIs are emerging as one of the more credible paths to get there.
Discussions at DTW will explore what it means to innovate with network APIs, how operators can leverage them alongside TM Forum Open APIs, and how the enterprise and platform developer communities should approach these capabilities for augmenting mobile applications and services.
See You in Copenhagen
AI adoption in telecom has crossed a threshold: it’s no longer about whether to deploy, but about how quickly operators can scale it into production without losing trust. As AI is simultaneously the biggest opportunity and the most complex operational challenge operators have faced in a generation, DTW 2026 is where the industry will come together to confront it.
As these themes gain momentum, RADCOM is helping build the assurance and intelligence foundations that make AI-driven operations possible. At DTW Ignite this year, we’ll be showcasing our High-Capacity User Analytics, which can cut operator TCO by up to 70% and accelerate the path to Level 4 autonomy through a closed-loop approach. We’ll also be demonstrating RADCOM Neura, our AI Agent Suite. Built on the RADCOM ACE platform, which correlates data end-to-end from RAN to core, Neura offers a scalable alternative to costly token-driven AI, grounded in trusted, real-time data and complete network visibility.
Whether you’re exploring the path to network autonomy, trying to scale AI beyond the pilot phase, or navigating the trust question AI raises, visit RADCOM at DTW Ignite where we will have the conversations that matter.
We look forward to seeing you there. Book a time to meet with us here: https://radcom.com/events/digital-transformation-world/
