Tech
Gartner: How AI will transform managed network services | Computer Weekly
In 2024, nearly all the service providers Gartner profiled in its Magic Quadrant for global WAN services report and the Magic Quadrant for managed network services report said they had started leveraging artificial intelligence (AI) in several ways to support the operation of enterprise networks. Areas of usage include AI for IT operations (AIOps), generative AI (GenAI) as a network assistant, enhanced service delivery, and AI in secure access service edge (SASE) and network security.
AIOps has emerged as a foundational capability in managed networking. Leading service providers, such as HCLTech, Microland and NTT Data, have begun to integrate AIOps capabilities and network automation for service onboarding and customer experience improvements. Also, service providers are deploying AI and/or machine learning (ML) to monitor network health, detect anomalies and automate routine tasks in network operations centres (NOCs).
The goal is to shift from reactive troubleshooting to proactive assurance. For example, if latency on a wide-area network (WAN) link starts spiking intermittently, a machine learning model might recognise the pattern as a precursor to link failure and alert engineers or trigger failover before a major outage occurs.
One such service provider is Tata Communications, which has invested in AI-based fault diagnosis using AI/ML for 85% accuracy, while AI-driven telemetry predicts and addresses issues for proactive network monitoring.
Also, many network equipment suppliers now embed AI features to support service providers for network monitoring.
GenAI as a network assistant
Over the past year, Gartner has seen a great deal of interest from managed network service (MNS) providers in applying GenAI to IT operations, including network management. The vision is to provide a network AI assistant that can interact with the provider’s operations teams via a natural language chat interface, help troubleshoot issues, document networks and even implement changes by generating configurations from intent.
One example is HCLTech, which is focusing on leveraging GenAI integrations with software-defined wide-area networking (SD-WAN) to deliver complete automation for lifecycle operations. It is building a supplier-focused GenAI large language model (LLM) as part of its service delivery platform (SDP).
Enhanced service delivery
AI is also leveraged in customer-facing aspects of MNS. Service providers are increasingly using AI to improve support and transparency for clients. This includes AI-powered customer service bots, service portals, and AI-generated reports or insights.
For example, many MNS providers profiled in the Gartner Magic Quadrant for managed network services report use bots, which are increasingly enhanced with AI capabilities, to automate repetitive tasks. Some have thousands of bots as part of their network automation codebases.
AI in SASE and network security
AI and ML are proving just as critical in the security side of MNS as they are in performance management. In fact, many service providers (for example, XTIUM and Microland) pitch AI-powered enhancements of their network security offerings, where the platform uses advanced analytics, AI and GenAI to strengthen and simplify management of local area network (LAN), WAN and cloud security.
For SASE and network security, AI can be used for automated anomaly detection. Here, the system quarantines a suspicious device or triggers multifactor authentication for a user behaving abnormally.
In policy optimisation, AI can recommend tightening or adjusting security policies, based on observed usage. For example, it can suggest zero-trust rules for an application, based on the context – location, time, company departments and so on.
Some advanced service providers, such as HCLTech, are exploring LLMs to assist security analysts – for example, summarising multistep attacks, or even writing firewall rules based on high-level descriptions of a threat.
Also, many SASE platform suppliers emphasise their AI/ML capabilities. For example, Versa Networks touts AI/ML-powered unified SASE that blends SD-WAN and cloud security, using ML to continuously adapt to network conditions and security threats. Similarly, Cato Networks highlights that it leverages AI/ML across its cloud-native SASE service to provide “reliable, accurate network security”, applying advanced data science to threat prevention and smart traffic management.
AI in MNS in 2028 and beyond
The integration of AI in MNS will increasingly enhance operational efficiency and enable more informed decision-making, ensuring that networks are robust and agile enough to adapt to changing demands and traffic patterns. Looking ahead three to five years from now, significant transformation in MNS is expected due to extensive use of AI – traditional, generative and agentic – and automation.
Widespread NOC assistants
The current rapid pace of development suggests that, by 2028, GenAI will have become a mature, trusted assistant in network operations. The experimental and nascent deployments of 2023 to 2024 will give way to robust network AI assistants embedded in MNS workflows.
These assistants will interface through natural language (text or voice) and be integrated with monitoring and ticketing systems. They will be able to answer complex queries about the network, draft change plans, and summarise incidents and problems.
Essentially, if 2023 was the introductory year for network AI assistants (see What is a network AI assistant?), by 2028, they will become a standard capability for NOCs to boost productivity.
The models behind the AI assistants are expected to be more specialised in network engineering and fine-tuned with each provider’s historical data, making them more accurate and context-aware than current tools are.
The best providers will leverage proprietary models – or at least proprietary fine-tuning – that become part of their intellectual property. For example, a provider can use a model trained on years of network event management data, which is exceptionally good at diagnosing telecoms network issues or in network security design efficacy. This will be a differentiator versus others that are using off-the-shelf network AI assistants.
By 2028, agentic AI will likely manifest as automated “Tier 0” responders in NOCs. These are AI agents capable of perceiving network incidents, understanding intent, making autonomous decisions, and executing actions for handling specific tasks and incident types end-to-end without human intervention.
By 2028, it is likely that many service providers will have enabled fully automated remediation for known issues. For example, if a branch SD-WAN router goes offline, the AI agent can perceive the incident, decide on a sequence of fixes – restart a virtual instance, fail over to backup, and so on – and execute them. It will alert a human only if those fail.
Another example could be the detection of a known bug, such as a memory leak in a firewall causing a slowdown. The AI agent, after perceiving the issue, will decide on a temporary configuration workaround or initiate a software patch, and execute these actions.
This goes beyond today’s static scripts by adding autonomous decision-making and action. The agent can verify if the issue truly matches a known pattern, using machine learning, and check if conditions are safe to execute the fix now, using policy – for example, it will reboot after business hours only if it is critical.
Fully autonomous networks will likely remain out of reach until well after 2028. But we expect that, by 2028, such self-healing actions will be accepted for narrow scopes, as service providers will have gained trust in AI for these repetitive tasks, thanks to long training and previous successful outcomes.
Nevertheless, the complexity of coordinating across domains means humans will still handle high-level decision-making. But for routine faults and performance tweaks, automated agents could become the norm, improving service reliability.
This article is based on an excerpt of Gartner’s AI will transform managed network services in the next three years report, by Gartner senior director analyst Gaspar Valdivia.