Your AI Journey: Finding the Use Cases That Matter
Improving operational efficiency with AI is non-negotiable in telecommunications. Discover how targeted AI use cases can deliver immediate value.
Analyst insight
10 December 2025
Author
Here is what the CCS Insight Predictions Event in London taught me. Read my reflections on the future of Network Assurance, AI and Data Management in Telecom.
Analyst insight
10 December, 2025
Author
I recently attended the CCS Insight Predictions 2026 event in London, surrounded by the usual mix of analysts, operators, vendors, and the inevitable person at the coffee station who seemed to have strong opinions about market consolidation. But underneath the noise, something important was emerging from the data: the operators who will lead the next decade are the ones making decisive moves on AI-driven network assurance right now.
Let me explain why.
The Prediction That Changed Everything
Among the hundred-plus predictions CCS Insight unveiled, one stood out to me: "In 2027, operators begin reporting financial and operational metrics to show the value of AI to their business."
On the surface, this seems obvious. Of course operators will measure AI value. But what this really signals is a market inflection point. For years, AI has been largely a "trust me, it's good for you" proposition. By 2027, that changes. The operators who've invested in AI assurance platforms now will have proven ROI to show their boards. The ones still deliberating? They'll be playing catch-up.
I watched a mid-market operator from central Europe in the corridor afterwards. She said something I won't forget: "We've spent three years planning the perfect AI roadmap. We should have just started."
That comment stayed with me through the rest of the day.
Agentic AI: Enterprise First, Consumer Later
Another prediction worth examining: "Agentic AI gains headway in businesses first, with consumer agents advancing little by 2028."
The room. The assumption has been that consumer AI would lead the pack. But the reality is simpler: enterprises need AI agents that work reliably within constrained environments. Network operations is exactly that kind of domain. When I heard this, I immediately thought of Polystar's agentic AI roadmap - autonomous agents that orchestrate network assurance without constant human intervention.
By 2028, the operators deploying these agents will have fundamentally transformed how they handle network complexity. The ones still managing networks reactively will wonder why they didn't move faster.
The Cloud Capital Inflection
Here's one that should worry anyone running infrastructure today: "Cloud capital spending continues to grow in 2026 before stalling abruptly in 2027."
This isn't saying cloud is dying. It's saying the window for establishing cloud-native competitive advantages closes in 2027. After that, you're adopting someone else's best practices, not leading them yourself.
For network assurance, this matters enormously. The operators who move to cloud-native assurance platforms in 2026 will have spent 2027 and 2028 optimizing operations, building institutional knowledge, and - critically - understanding how to extract maximum value from their cloud deployments. The operators who move in 2028 will be scrambling just to keep up.
The irony? Many mid-market operators still think they have time to plan this transition. They don't.
Specialized AI Beats General Purpose
CCS Insight predicts that "by 2027, small, focused AI models process more output tokens than large, general-purpose models."
This is profound. The entire industry has been obsessed with scale - the largest language models, the most parameters, the highest processing capacity. But the prediction suggests we're about to pivot toward specialization.
In network assurance, this is already happening. You don't need a general-purpose model trying to solve every problem in telecom. You need purpose-built models designed specifically for anomaly detection, root-cause analysis, and service quality prediction. Models that understand network topology, traffic patterns, and customer behavior in ways a general-purpose system never could.
This validates what I've been seeing in the market: vendors who've built domain-specific AI are outperforming those trying to force general-purpose models into specialized use cases.
The Practical AI that Actually Gets Used
One more prediction that felt honest: "By 2028, automatic summarization, voice-to-text transcription, image editing and live translation are the most-used generative AI features."
Notice what's not on that list? The flashy stuff. The revolutionary breakthroughs. The things that sound impressive in earnings calls.
What's on the list are practical tools that solve real problems. Operators don't want AI that's intellectually interesting. They want AI that helps their engineers troubleshoot faster, that summarizes incidents automatically, that surfaces the information they need without requiring them to dig through data for an hour.
This matters because it reveals what will actually drive adoption. The assurance platforms that win won't be the ones with the most advanced AI. They'll be the ones that make operators' daily work easier.
Data Partnerships Are the New Moat
Here's a prediction that caught me off guard: "By 2029, the value of AI platforms is increasingly derived from their content agreements as platforms become commoditized."
As assurance platforms begin to look more similar to each other, the real differentiation will come from data - who you have partnerships with, what data sources you can access, how you can correlate insights across different domains and vendors.
This suggests that the operators who build strong partnerships around data - who can ingest performance data from multiple vendors, correlate it intelligently, and derive insights that single-vendor platforms can't - will have defensible competitive advantages that pure software capability alone won't provide.
Sovereignty and Resilience Aren't Optional
Two predictions crystallized something I've been thinking about: "By 2027, EU member states mandate satellite backup for critical national infrastructure" and "By 2028, the US Department of Defense selects Samsung to build private 5G networks at its Indo-Pacific bases."
These aren't just regulatory headlines. They're signals that governments are taking control of network resilience seriously. For mid-market operators, this means the assurance systems you deploy need to work across vendor landscapes you don't control, in geographies with different regulatory frameworks, in scenarios where uptime isn't negotiable.
Vendor-independent network assurance isn't a nice-to-have anymore. It's becoming table stakes.
The Human-AI Partnership That Actually Works
Perhaps my favorite pair of predictions: operators will realize that replacing entry-level employees with AI was a mistake by 2030, while simultaneously reporting strong financial returns from AI deployment by 2027.
These aren't contradictory. They reveal the truth: AI augments human expertise. It doesn't replace it. The operators who'll win are the ones who use assurance platforms to make their best engineers more effective, not to eliminate the need for skilled teams.
Your network assurance platform should make an expert engineer capable of managing 10x the complexity, not attempt to replace five engineers with a bot.
What This Means for Your 2026 Roadmap
If you're a mid-market operator, the CCS Insight predictions point to a clear window: 2026 to 2027.
This is when you need to have made decisive moves on cloud-native assurance platforms. This is when you need to have started deploying agentic AI into your operations. This is when you need to be building partnerships around data, not just buying software.
By 2028, the leaders will be optimizing and scaling. By 2029, the laggards will be explaining why they're still managing networks the old way.
The operators I spoke to in London who "get it" all said the same thing: they're not waiting for perfection. They're moving now. They're learning as they go. They're accepting that the platform you deploy in 2026 won't be the same as the one you're running in 2028 - and that's okay. Staying ahead matters more than waiting for the perfect solution.
The Uncomfortable Truth
Here's what struck me leaving London: the operators who will struggle most aren't the ones without the budget for new assurance platforms. They're the ones who convinced themselves they had time to plan.
Time is the scarcest resource in 2026. Not money. Not technology. Time.
The window to establish competitive advantages through AI-driven, cloud-native, vendor-independent network assurance is open now. It starts closing in 2027. By 2029, it's effectively shut.
The operators making moves today - even imperfect moves - will be leading their markets through 2030 and beyond.
The operators still planning? They'll be wondering where the time went.
Elisa Industriq SVP Ian Teasdale is a telecoms veteran focused on network operations, automation, and service assurance. He attended the CCS Insight Predictions event in London in October 2025.
Polystar is an Elisa Industriq company providing commercial-grade network assurance and service assurance solutions to over 100 CSPs globally.
