By Viet Nguyen, President, 5G Americas (October 2025) –
Imagine standing at a busy intersection with traffic lights running on fixed timers. They change on schedule whether or not cars are actually waiting. Now imagine a smart intersection where sensors adjust the lights based on real traffic flow, keeping everyone moving efficiently. This is the difference between today’s automated networks and tomorrow’s intent-based autonomous networks (IBANs). Instead of just following pre-set rules, IBANs understand the goal—smooth traffic—and adjust in real time to get there.
In 5G Americas’ white paper “Enabling Intent-Based Autonomous Networks,” we explore how telecom operators are increasingly harnessing the power of AI-driven, self-configuring, and self-healing systems to meet the rising complexity of 5G, AI, IoT, and cloud-native services.
As wireless networks scale in complexity with more services, more devices, wider spectrum channels, and varied user expectations, it is no longer sufficient to automate tasks. We need systems that understand not just what needs to happen but why. IBANs give operators the capability to express outcomes such as performance, reliability, and regulatory compliance, while letting the network decide how best to deliver them. This is no longer just a vision; elements of this shift are already appearing in live networks.
So let’s understand how that works. Intent works by separating goals from execution. Rather than configuring every rule and path manually, operators specify desired behaviors, such as maximum latency for VR applications or keeping sensitive traffic within certain jurisdictions, and the network carries out appropriate actions. Think of it like telling a navigation app your destination rather than dictating every turn. The system monitors its own performance through telemetry and KPIs, much like the app checking real-time traffic. When conditions change from demand spikes, interference, or policy updates, the network reroutes itself to stay aligned with the operator’s goals.
A key feature is persistence: once set, an intent remains active until revoked. This means continuous alignment with business or operator objectives from service level agreements (SLAs) to regulatory mandates. This ongoing oversight is already evident in trials and early implementations. We see pieces of IBAN logic in use today through fault-prediction systems, traffic forecasting, and AI-assisted network orchestration.
Real-world examples of intent-based automation are already proving valuable across several operational domains. One prominent example is customer experience assurance. In this case, intent-driven systems translate high-level goals—such as maintaining video call quality or reducing gaming latency—into concrete actions by continuously monitoring KPIs and adapting network parameters like throughput, jitter, or path selection. This transformation from reactive troubleshooting to proactive experience management significantly reduces churn and enhances user satisfaction.
Reasoning and utility models allow networks to balance competing priorities. Imagine a power grid deciding where to send electricity first during a peak surge—hospitals get priority, then homes, then factories. IBANs work the same way for data traffic, choosing in real time whether to reroute traffic, allocate spectrum, or dial back less-critical services. Over time these systems become better at predicting which choices will satisfy overall goals.
Additionally, intent applies across multiple abstraction layers. At the top, business intents express strategic aims such as compliance, customer satisfaction, and sustainability. Service-level intents define metrics like latency, throughput, and availability. Resource-level intents translate those into configuration decisions: which functions to deploy, what slices are needed, and how infrastructure should flex. You can think of this like city planning: leaders decide on the vision (business intents), engineers design road capacity and traffic rules (service intents), and construction teams lay down the asphalt and signals (resource intents).
There are also additional “non-functional” requirements that are increasingly important. Security, data locality, privacy, and environmental constraints are now expected in network contracts, regulation, and user expectations. An intent may require that customer data remain in a specific region or that the network satisfies encryption or data retention policies. The network must enforce these automatically, not simply document them.
IBANs also improve risk management. Networks should act autonomously when risk is low and escalate when risk is high. This is similar to an airplane’s autopilot: it can handle routine turbulence on its own, but calls for the pilot if conditions exceed safety limits. Clear feedback, human-in-loop options, and mechanisms to override automated decisions will be essential.
Of course, there are also challenges. Observability must be precise, with metrics and telemetry that are granular and timely. Designing utility models that correctly represent operator priorities without unintended consequences is complex. Managing non-functional intents like security or privacy requires legal, engineering, and policy coordination. Legacy infrastructure may also slow adoption in some regions.
Moreover, as networks become increasingly software-defined and cloud-native, open APIs and observability frameworks become essential for intent-based operations to function at scale. Open APIs ensure interoperability across multi-vendor environments by separating the “why” of business intent from the “how” of execution. For instance, a CSP may define an intent to minimize latency for augmented reality services, while underlying systems autonomously reconfigure routing, compute placement, and bandwidth allocation. Observability frameworks support this closed-loop automation by providing real-time visibility into system behavior, enabling AI models to detect anomalies, predict future states, and trigger corrective actions.
Therefore, standardization and common models are critical for scaling IBANs. Projects such as the TM Forum’s Intent Ontology (TIO) and Autonomous Network Level Evaluations Tool (ANLET) guide operators in benchmarking their progress and defining implementation roadmaps and provide a shared vocabulary and process. Without compatible semantics, intent definitions risk diverging, frustrating interoperability. It’s like having road signs that look different in every city—drivers would be confused, and traffic would stall. Common standards give everyone the same “language” for navigation.
The importance of IBANs becomes even clearer as 5G-Advanced capabilities roll out and work toward 6G begins. As networks become more software defined, cloud native, and AI-enabled, many of the foundational elements needed for intent-based autonomy are being deployed now. Ultra-reliable low latency services, dynamic slicing, and adaptive beamforming will all benefit from IBAN adoption.
From my perspective at 5G Americas, the industry is at a decisive moment. The decisions we make about intent common models, lifecycle frameworks, human oversight, and vendor interoperability will shape networks over the next few years. My belief is that those who adopt IBANs thoughtfully will achieve not just efficiency but also resilience and strategic flexibility.
Operators, vendors, standards bodies, and policymakers all have roles to play. This is the time to align on intent frameworks, demand interoperability, pilot aggressively, and ensure transparency in automation. The future of wireless is not just faster networks; it is networks that understand what is required of them and adapt when those requirements change.
—Viet
