Autonomous networks have become a defining ambition for the telecom industry. The telecom and network management market for agentic AI is projected to nearly triple over the next five years, reaching an estimated $11.09 billion by 2030.
The biggest misconception telecom leaders have about agentic AI is that it’s primarily a technology challenge. In reality, success depends on trusted data, well-defined operational processes, governance, systems integration, and clearly defined decision boundaries for AI agents.
Yet most of the conversation still focuses on what AI can do, from detecting network faults to automating remediation. Far less attention is being paid to a more fundamental question: Do telecom operators have the operational foundations to trust AI with decisions that directly affect network performance, service reliability, and customer experience?
The foundation AI agents depend on
Before agentic AI can operate safely and reliably at scale, operators will need to establish the datasets, guardrails, and workflows that allow agents to be effective.
Today, many telecom operators add AI agents directly to legacy OSS/BSS environments. But when operational data, network inventory, service information, and customer impact data remain siloed, those agents can optimize only within individual systems — not across the network as a whole.
This is often where agentic AI initiatives stall. A pilot can succeed in a controlled environment and still lose momentum once it needs to operate reliably in production, across live systems and real operational risk.
Even when organizations manage to overcome the data and integration challenges, many still haven’t established the clear decision boundaries, oversight, and accountability to fully trust the decisions and actions AI agents take.
A recent survey of global telecom operators by TM Forum found that just 14% of operators can produce externally reviewable evidence demonstrating their AI systems are trustworthy. When telecoms can’t trust their AI agents to make the right calls, employees remain the final decision-makers, forcing AI into an advisory role instead of the operational role it was intended to play.
This dynamic plays out in network operations centers, where a single network fault can trigger thousands of alerts across multiple systems, leaving engineers to correlate events and identify the root cause.
An AI agent could automatically correlate those alerts, identify the underlying issue, assess customer impact, and even initiate approved remediation. But telecoms only realize those gains when they establish the governance to define what actions AI can take, under what conditions, and with what oversight. Otherwise, every recommendation still requires human review, slowing resolution instead of accelerating it.
4 steps for building AI-ready network operations
Realizing the value of agentic AI depends less on the AI itself than on the environment around it. For telecom operators, that requires strengthening the data, governance, and operational foundations that allow AI to scale safely and reliably. The following four steps can help operators lay the groundwork.
1. Start with a few high-value use cases
The vision of fully autonomous network operations can make it tempting to automate multiple workflows at once. But broad deployments expose gaps in data, governance, and operational processes that become much harder to fix at scale.
Instead, start with a select handful of high-value use cases where agentic AI can deliver measurable value with manageable risk, such as alert prioritization, cross-domain incident correlation, or root cause analysis. Focused deployments let operators validate AI performance and uncover foundational gaps early, strengthening their operating environment before expanding AI into more complex network operations.
2. Establish governance in advance
As AI takes on more operational responsibility, operators first need to decide where its authority begins and ends. Clearly defining which decisions agents can make independently, which require human involvement, and how every action will be monitored creates the accountability needed to confidently deploy AI at scale. That kind of governance rarely happens on its own. It takes executive sponsorship to prioritize the work, standardized APIs to connect systems consistently, and observability to track what agents are actually doing across the environment.
Proactively establishing those guardrails allows organizations to expand AI’s role over time while ensuring people remain responsible for policy, regulatory compliance, and other high-impact decisions.
The distinction that separates fast-moving operators from slow ones is where governance enters the process. When legal and security are part of the conversation from the start of an idea, teams don’t waste cycles building something that can’t be deployed — and they don’t accumulate the trust debt that slows every subsequent decision.
3. Build a trusted data foundation
AI can only make reliable decisions if it has the right operational context. In many telecom environments, data remains spread across network, IT, and service domains, preventing AI from seeing the complete picture.
Consider what this looks like in practice. A leading US 5G operator needed to manage network capacity during a major national public event. By unifying real-time network traffic data, capacity benchmarks, and field operations workflows on a shared data platform, their team was able to identify potential congestion along the event route several hours in advance and make targeted capacity adjustments before any customer experienced an issue. Without that unified view — where network load, coverage gaps, and field response systems are all visible in one place — an AI agent has no basis for a proactive recommendation. It can only react after the problem surfaces.
Before AI can reliably support network operations, operators need trusted, governed data that’s securely available in near real time, along with a common semantic understanding across systems so AI interprets network events consistently. Modern data platforms help create that foundation by connecting fragmented operational data into a unified view while maintaining governance and security.
4. Scale based on measurable outcomes
Once organizations see early success, the instinct is often to expand AI as quickly as possible. But scaling autonomy before operational processes are ready can recreate the same challenges that slowed adoption in the first place.
Each deployment should earn the next. As AI delivers measurable improvements in incident response, operational efficiency, or network resilience, operators can confidently expand its role while ensuring the supporting foundation continues to mature alongside it.
Establishing trust in autonomous networks
AI capabilities have advanced rapidly across telecom. The industry’s next challenge is making sure operators have the systems, data, and processes to use them effectively.
Building trust in agentic AI requires telecom operators to invest as seriously in data quality, governance, and operational maturity as they do in the technology itself. Only then can autonomous networks become an operational reality rather than a long-term industry goal. The objective isn’t to remove people from network operations. It’s to put them in a position to supervise increasingly autonomous systems and keep improving them.
The views expressed in this article belong solely to the author and do not represent The Fast Mode. While information provided in this post is obtained from sources believed by The Fast Mode to be reliable, The Fast Mode is not liable for any losses or damages arising from any information limitations, changes, inaccuracies, misrepresentations, omissions or errors contained therein. The heading is for ease of reference and shall not be deemed to influence the information presented.