The Operational Gap That's Stalling Autonomous Networking
Autonomous networking requires that every action be verified against a mathematically accurate model of the full production network before execution.
Nikhil Handigol, Forbes Councils Member
Forbes
3 min read
6/10
Key Takeaways
Only 12% of enterprises have deployed autonomous network operations beyond isolated use cases, per a 2025 Gartner survey.
Network outages caused by automation errors remain a top concern; 43% of operators cite verification gaps as the primary barrier to automation (Internet Society study).
Large enterprises lose an average of $300,000 per hour of network downtime (IDC), underscoring the cost of the operational gap.
Vendors such as Forward Networks, Itential, and Nokia are developing digital twin solutions to provide real-time network state models for pre-change verification.
The operational gap between intended and actual network state is the biggest obstacle to achieving intent-based networking (IBN) at scale.
Despite bold promises from vendors, the march toward fully autonomous networks has hit a wall—and the culprit is a stubborn operational gap that no amount of AI hype can bridge. Autonomous networking requires that every action—every configuration change, every routing update—be verified against a mathematically accurate model of the full production network before execution. That ideal remains out of reach for most organizations today. Network engineers at enterprises and service providers alike are discovering that existing automation tools lack the real-time, high-fidelity digital twins needed to validate changes safely at scale, leaving the industry stuck between legacy manual processes and true zero-touch operations. The problem is not a lack of intent. For years, the networking industry has chased intent-based networking (IBN), where operators declare what they want and the network figures out how to deliver it. Vendors from Cisco to Juniper have rolled out controllers and orchestration platforms. Yet the operational gap—the disconnect between the network's intended state and its actual, dynamic state—persists. Every network is a living system: cables get unplugged, traffic spikes, misconfigurations creep in. A static model cannot capture that. Autonomous networking demands continuous, mathematically rigorous verification, not occasional audits. The core issue is that most organizations today operate their networks using configuration management tools that assume a stable, predictable environment. But production networks are chaotic. When a change is pushed—say, a new ACL or BGP policy—engineers often rely on pre-change simulations that quickly become stale. The network's actual state may differ from the model due to unlogged modifications, hardware failures, or third-party integrations. Without a live, always-updated digital twin that mirrors the exact state of every switch, router, and firewall, automated actions risk disrupting services. Industry research underscores the scale of the challenge. A 2025 Gartner survey found that only 12% of enterprises had deployed any form of autonomous network operation beyond isolated use cases. Meanwhile, a study by the Internet Society highlighted that network outages caused by automation errors remain a top concern, with 43% of operators citing verification gaps as the primary barrier to further automation. Companies like Forward Networks, Itential, and Nokia are pushing digital twin solutions that promise real-time network modeling, but adoption remains slow. The operational gap is not just a technical hurdle—it has financial and security implications. For every hour a network is down, large enterprises lose an average of $300,000, according to an IDC analysis. In sectors like finance or healthcare, the cost is even higher. The inability to trust automated changes forces teams to keep humans in the loop, undermining the ‘autonomous’ label. As one network architect at a Fortune 500 bank put it (off the record): 'We have automated everything we can—but we still watch the changes with our own eyes.' The industry is now pivoting toward 'continuous verification' approaches, where network models are updated in near real-time using telemetry data. Startups are developing AI that can detect drift between intent and reality in milliseconds. The operational gap will likely narrow as these technologies mature, but true autonomy may remain a five-year horizon. The key milestones to watch are the integration of real-time telemetry with digital twins and the emergence of open standards for model accuracy. For now, the operational gap is the silent saboteur of autonomous networking—and closing it is the single most important task for the next generation of network engineers.
Frequently Asked Questions
The operational gap is the disconnect between a network's intended state and its actual, live state. Autonomous networking requires that every action be verified against an accurate model of the production network, but constant changes and unlogged modifications make it difficult to maintain such a model in real time, creating a gap that stalls full automation.
Autonomous networking is stalling because existing automation tools lack high-fidelity, real-time digital twins of the full production network. Without continuous verification against an accurate model, automated changes risk outages, forcing operators to keep humans in the loop and preventing true zero-touch operations.
A network digital twin is a virtual replica of the physical network that mirrors its state in near real-time using telemetry data. It allows engineers to simulate and verify changes before deployment, helping close the operational gap necessary for autonomous networking.
Organizations can close the operational gap by adopting continuous verification platforms that update network models in real time, integrating AI to detect drift between intent and actual state, and investing in digital twin technology from vendors like Forward Networks and Itential.
Failing to address the operational gap means continued reliance on manual processes, higher risk of misconfigurations and outages, and inability to achieve network automation at scale. This leads to increased operational costs, slower service delivery, and potential revenue loss from downtime.
Intent-based networking is an approach where operators declare high-level business or operational intent, and the network automatically translates that into configuration changes. However, IBN has been limited by the operational gap—without accurate real-time models, the network cannot reliably execute intent without human oversight.
