Report: Key challenges before broadband networks are autonomous
October 8, 2026



Broadband service providers are readying their networks for AI-driven autonomy in order to deliver better broadband services. However, technical and operational hurdles remain in their way, according to a Broadband Forum report. The research highlights poor device compatibility and inconsistent data as key challenges that must be addressed before broadband networks can become autonomous.

The report reveals strong commitment to autonomous networking, with 94 per cent of service providers planning to use software-defined networking (SDN) as part of their autonomous networking journey and 92 per cent describing AI/ML integration as critical or important to these strategies. However, poor and inconsistent data from multiple systems was identified by 83 per cent of service providers as a major barrier to deploying AI successfully in their networks.


“A software‑driven foundation gives networks the ability to execute an instruction on its own, while AI provides the intelligence to let it make decisions and adapt to changing conditions automatically,” said Broadband Forum CEO, Craig Thomas.

“Broadband networks are increasingly assembled from hardware, software, cloud platforms and management systems supplied by different vendors, while automation relies on data and control interfaces that need to work consistently across them. Standards make this complex environment easier to integrate and operate at scale by reducing the technical and operational friction that makes those changes harder to implement,” Thomas continued.

Service providers said that predictive maintenance was the main area where they were or were planning to use AI and ML (90 per cent), followed by anomaly detection (72 per cent) and automated fault resolution (69 per cent). The main use cases for AI deployment were customer experience optimization (87 per cent) and service assurance (85 per cent).

“AI needs to understand what that data means and how everything is connected before it can make the right automated decisions,” said Incognito Software Systems Head of Product, Craig Sinasac. “Standards support that evolution by creating consistent definitions, identifiers, and models that describe how network resources relate to the services they deliver and the events that affect them,” he commented.

Device compatibility also remains a major hurdle in the adoption of a software-driven foundation that AI and autonomy depend on, with 78 per cent of respondents citing poor interoperability between vendors as a major barrier. This is further complicated by the need to integrate new SDN platforms with legacy infrastructure, including DSL, cable/HFC and multiple generations of fiber equipment. As a result, 90 per cent of respondents identify integration with existing systems as a key challenge.

“While standards can reduce the friction of integrating and scaling automation for service providers, their involvement must go beyond writing specifications,” added Radisys Senior Product Manager Connect Modular Broadband, Robert Soukup. “Taking part in reference implementations, interoperability testing, and proof-of-concept programs enables service providers and vendors to validate common approaches against real network lifecycles and feedback practical deployment experience back into the standard.”