- An Accenture survey found 79% of telcos are still at Level 0/Level 1 autonomy, with only 22% expecting to reach Level 4 automation by 2030
- Key obstacles include legacy BSS/OSS systems and limited talent, leaving many operators stuck in an incremental, hybrid approach to automation
- Hyperscalers are ahead in automation maturity but they struggle with the same network resiliency issues as telcos
Achieving network autonomy is a top priority for telcos in their quest for network efficiency and AI enablement, but most have barely scratched the surface, according to recent Accenture research.
A survey of “a few hundred” telecom executives found roughly 79% of operators are still in the Level 0/Level 1 phase, “fairly manual” with some level of automation and assisted operations capabilities, said Tejas Rao, managing director of Accenture’s Communications, Media and High-Tech practice.
Respondents are cautious about reaching full autonomy, as only 22% expect to reach Level 4 automation by 2030 and just 10% think they’ll get to Level 5 by that time.
There are a few outlier operators that have already rolled out Level 4 networks, such as China Mobile, Malaysia’s DNB and Vivo in Brazil. But the problem for many telcos, Rao told Fierce, is that they’re “stuck in this incremental hybrid approach.”
Instead of reimagining processes from the ground-up to support automation, he said telcos typically work “on a few use cases here and there,” integrating AI into existing operations for incremental improvements.
Part of the challenge lies in aging BSS/OSS platforms and technical debt, with 47% of Accenture respondents saying legacy platforms limit interoperability. Whereas 49% cited limited talent in AI and automation engineering.
Another issue is figuring out how to shift from using AI reactively to predicting incidents and network failures before they happen – a feat easier said than done.
At Mobile World Congress 2026 in Barcelona, T-Mobile CTO John Saw said the operator is already using some autonomous network technology, notably to make about 30,000 antenna tweaks to keep customers connected during Winter Storm Fern.
For Accenture’s part, the company in February acquired an “advanced AI” solution from Avanseus that provides telcos models for prediction, anomaly detection and optimization for complex network operations.
Telcos look to hyperscalers for autonomous advice
Avanseus’ tech is also designed for seamless integration across hyperscaler agentic AI platforms, with Rao pointing out telcos are taking inspiration from hyperscalers in writing their autonomy playbook.
“We’re helping [to] kind of bridge the gap between how it’s being done today in that industry and what the applicability could be for telcos themselves,” he said. “I think the tooling is there, right, the adoption is there in terms of understanding.”
AWS’s fiber network is largely autonomous, the hyperscaler told Fierce in January, and Google Cloud is also striving for full automation. But while hyperscalers are ahead of the curve on autonomy compared to telcos, they’re not immune to outages.
Digital twin’s potential for network autonomy
Digital twins – digital replicas of physical or virtual networks – could offer telcos “a huge opportunity” in designing autonomous network layouts, Rao said.
The technology has taken longer than expected to take off due to high maintenance costs, but AI could help pick up adoption. Vendors like Forward Networks and Viavi have both integrated AI capabilities in their digital twin technologies.
“I kind of think of digital twins as looking at various lifecycle phases of a network operator,” said Rao. As operators increase their capital spend on infrastructure upgrades, they can perform an “ongoing simulation, not just a one-time simulation but using it as a dynamic platform capability” that can help them decide where to move their capital in real time.
Read all of our coverage from Mobile World Congress 2026 in Barcelona here.