Telcos Face Software, Governance Gaps on Path to Network Autonomy

The automation-autonomy divide

Telecommunications operators pursuing autonomous networks face significant gaps in software deployment practices and AI governance, according to vendor and carrier executives speaking at DTW Ignite 2025 in Copenhagen.

The distinction between automation and autonomy is more than semantic. Anders Vestergren, head of Solution Area Network Management at Ericsson, compared automation to a car with an automatic transmission—it simplifies one task. True autonomy, he said, resembles a self-driving vehicle that understands intent and executes all necessary steps to reach a destination.

The path to autonomy varies dramatically by use case. Vestergren noted that achieving Level 4 autonomy in energy management can be relatively straightforward because it operates within a single domain using mature autonomous features. Other use cases present far greater challenges.

Process versus domain autonomy

Not all operators agree that full domain-level autonomy should be the primary goal. T-Mobile Chief Network Officer Ankur Kapoor told the conference his team focuses on autonomous processes that deliver specific customer outcomes rather than vertical domain autonomy.

"If I can do a lot of autonomous operations in the core, that's a win, right? That helps a lot with network operations, but it doesn't help much with overall end-to-end service experience," Kapoor said.

He argued the industry should shift from traditional metrics like coverage, speed and latency toward actual customer outcomes. Customers care that services work, not how they're delivered.

Software deployment remains a barrier

Jan Hofmeyr, AWS vice president and telco chief, identified fundamental software deployment capabilities as a critical missing piece. He said operators need the ability to deploy code monthly at minimum and respond to zero-day vulnerabilities within the same day.

"I think there's a lot of work to be done to get the network to a point where just the basics around software deployments can be done," Hofmeyr said.

While many operators believe their software deployment practices are adequate, Hofmeyr noted that code deployments remain far from touchless. The human-in-the-loop approach persists across the industry.

AI governance requirements

Moving toward autonomous operations will require declarative guidelines defining what AI agents can and cannot do. Hofmeyr said this extends beyond traditional governance to include specific guardrails around destructive actions.

Operators won't allow AI agents to delete database entries anytime soon, he noted. The same checks and balances humans follow must apply to AI systems—but those frameworks don't yet exist.

Hofmeyr identified organizational factors as the primary obstacle, with people, workflows and process adaptation creating larger barriers than technology itself. In 70% of cases, he said, technology serves as a scapegoat for these human and process challenges.

Why it matters

The gap between automation and autonomy has direct business implications as operators invest heavily in AI-driven network operations. Without foundational software deployment capabilities and AI governance frameworks, carriers risk falling behind on innovation while becoming vulnerable to rapidly evolving security threats. The distinction also matters for vendor selection and technology roadmaps—solutions that automate individual tasks won't deliver the intent-driven, outcome-focused operations that define true network autonomy.

These details were first reported by Fierce Wireless from DTW Ignite 2025.