Enterprise AI is changing how organisations evaluate WAN and connectivity strategy. A Gartner® report described Lumen as “the company to beat” in this category.¹
The Gartner® report states: “Enterprise AI is redefining networking and connectivity services, shifting value from bandwidth to latency determinism, edge proximity and service integration. Lumen, with its unified service platform, is the company to beat in enterprise AI WAN and connectivity services.”
Gartner, AI Vendor Race: Lumen Is the Company to Beat in Enterprise WAN and Connectivity Services for AI, By Kameron Chao, Susan Welsh, etc., June 2026
From our perspective, these trends reinforce the importance of predictable performance, programmability and integrated service delivery.
But the more interesting question is a step behind it: why is enterprise connectivity suddenly being judged differently?
From bandwidth to predictable performance
For years, network planning has often started with two familiar questions: where do we need coverage, and how much capacity do we need? Those questions still matter. AI does not make them irrelevant. It simply exposes what they do not fully answer.
Training and inference can involve high-volume data movement across distributed environments. In that model, the question is no longer only whether there is enough capacity. It is whether performance can remain consistent across the route.
That is where predictability becomes strategically relevant. For latency-sensitive AI workloads, variable performance can affect how consistently services operate. Depending on the service, network design and applicable service commitments, enterprises can assess whether performance can be engineered and managed to meet defined requirements.
The shift, then, is not simply towards faster networks. It is towards networks that can give enterprises greater control over traffic paths and more consistency as demand changes.
Why programmability matters
The next shift is easy to miss. AI environments may not stay in one fixed shape.
A workload needs to run across enterprise infrastructure, cloud platforms or edge locations depending on the use case. As that happens, policies, capacity and routing decisions also need to change.
That can be difficult to manage when the network is built around fixed configurations and separate management layers. Programmable control can support a more flexible operating model. Depending on the service design and implementation, it can support on-demand connectivity provisioning, centralised policy application and improved visibility across enterprise, cloud and edge environments.
Network infrastructure, together with a cloud-native control plane, is designed to turn connectivity from a largely static design into a more dynamic operating environment.
At Lumen, we view an AI-ready network as combining network infrastructure, programmable connectivity and an ecosystem spanning cloud, edge and AI. Subject to service availability, design and implementation, these capabilities can help enterprises adapt connectivity as AI requirements change and reduce the need to manage each adjustment as a separate infrastructure project.
From separate components to an integrated service
AI also tests how well the operating model holds together.
Data, compute, cloud, network and security considerations do not sit neatly apart. When each part is managed as an isolated service, the organisation is left to stitch together the overall experience and resolve issues across multiple providers and control points.
An integrated service model can bring WAN, cloud and edge connectivity, orchestration, automation and service performance into a more cohesive environment. Subject to the services selected and their implementation, this approach can provide a more coordinated way to manage connectivity across relevant environments rather than treating each connection as an isolated component.
An integrated service model can support clearer operational responsibility. Enterprises can gain improved visibility into how performance is managed and where responsibility sits within the scope of the services provided, particularly when connectivity, infrastructure and security capabilities need to work together.
The network is becoming part of the AI architecture
This leads to the larger point. Enterprise AI architecture does not stop at models, applications and compute. The network influences how data reaches those environments, how workloads connect across them and whether the experience can be delivered consistently to users.
That is why WAN and connectivity strategy should be part of the AI planning conversation. Network leaders may need to consider where AI workloads will be located, how data will move between environments, what level of performance will be required and how policies will be applied across the full path.
Seen this way, the network is more than a transport layer. It can form part of the service platform connecting enterprise, cloud and edge environments, with control and visibility designed to support distributed AI workloads.
What does this mean for APAC enterprises?
For APAC enterprises, this reframes the assessment. Network footprint remains relevant, but organisations should also consider whether connectivity can be orchestrated across enterprise, cloud and edge environments with appropriate policy consistency, visibility and accountability, subject to local requirements, service availability and the capabilities of relevant providers.
That is a different way to think about AI connectivity: not as isolated country or provider decisions, but as a coordinated architecture designed around where data, workloads and users need to connect. For cross-border deployments, organisations should also assess applicable data protection, cybersecurity, data localisation, sector-specific and transfer requirements in each relevant jurisdiction. Network architecture can support an organisation’s compliance approach, but it does not by itself establish regulatory compliance.
A different way to evaluate enterprise connectivity
For enterprise leaders, evaluating WAN and connectivity strategy now requires a broader set of questions:
- Can the network deliver predictable performance for distributed AI workloads?
- Can it adapt as workload locations and requirements change?
- Can services be managed through a cohesive operating model rather than as separate components?
- And can the organisation maintain clear accountability across the full experience?
AI is not making the enterprise network less important. It is revealing how central the network already is.
Evaluate your AI connectivity strategy today
As AI workloads become more distributed, enterprises should assess whether their WAN and connectivity strategy is designed to support the performance, visibility and agility required for their specific workloads and operating environments.
Talk to a Lumen expert about enabling and orchestrating connectivity for the AI era.
¹Gartner, AI Vendor Race: Lumen Is the Company to Beat in Enterprise WAN and Connectivity Services for AI, By Kameron Chao, Susan Welsh, etc., June 2026
Gartner is a trademark of Gartner, Inc. and/or its affiliates.
Gartner does not endorse any company, vendor, product or service depicted in its publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner’s business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this publication, including any warranties of merchantability or fitness for a particular purpose.
Services not available everywhere. Business customers only. Lumen may change, cancel or substitute products and services, or vary them by service area at its sole discretion without notice. © 2026 Lumen Technologies. All rights reserved.
