A guest blogpost by Roy Chua, Founder and Principal at  AvidThink

Introduction

When we talk about convergence in enterprise networking, we generally refer to the convergence of connectivity and security. There is, however, another axis of convergence in enterprise networks — the convergence of multiple network domains (LAN, WAN, etc).

The main driver of convergence across both axes is IT leaders’ desire for simplicity and security. Their goal has been, and continues to be, a single unified management plane. This single pane of glass is meant to streamline operations while improving reliability and security.

How close are we to achieving convergence across two axes? What barriers remain? And how is AI changing the landscape? Read on to find out.

Enterprise Connectivity and Convergence

Convergence isn’t a new trend in enterprise networking. We’ve been hearing about “convergence” for well over a decade — arguably two. However, the convergence of networking and security, and the desire to have all networks be secure by default, is yet to be achieved. How do we know this?

When AvidThink ran the survey for the 2026 Enterprise Connectivity Report, only 35% of the organizations surveyed(1) indicated they had converged security with networking. That means most enterprises remain siloed.

The Other (Less Talked About) Convergence

That same survey surfaced another axis of convergence in enterprise networks. 80% of those surveyed want to manage multiple network domains in an integrated way, from campus Wi-Fi and wired Ethernet to WAN, SD-WAN, and SASE (which itself is a converged kitchen sink). To address customer preference for cross-domain convergence, leading enterprise networking vendors have created consolidated management platforms:

• Cisco: first unifying under the Meraki umbrella and then launching Cisco Cloud Control at Cisco Live US 2026
• Arista: purchasing the VeloCloud SD-WAN assets from Broadcom (which had acquired VMware), and working on unifying management (analytics first),
• Extreme Networks: launching their converged Platform ONE that manages their diverse product portfolio comprising acquisitions from Wi-Fi to wired Ethernet to SD-WAN,
• HPE: acquiring Juniper and then consolidating functions under the Mist and Aruba Central platforms, easing its path into cross-product management.

Additionally, the survey identified private 4G/5G as a desired add-on domain, and vendors like HPE and Cisco continue to dip their toes into the private mobile market.

This cross-domain convergence crosses multiple underlay fabrics. From wired Ethernet connections to Wi-Fi to SD-WAN over direct internet connections, broadband, MPLS, satellite, and 4G/5G fixed wireless access links, the variety to be handled and the complexity continue to grow.

What We Gain (and Lose) When the Walls Come Down

The survey showed that the main driver for convergence (across both axes) is the desire for a single unified management plane (80% of survey takers indicated this). The goal of enterprises is to streamline and reduce operational expenses and complexity that comes with a single pane of glass.

However, this single pane is only as effective as the data that feeds it. And corralling multiple per-domain tools within a single interface doesn’t increase visibility. Or put another way, convergence fixes fragmentation, not visibility.

Existing blind spots, ranging from opaque encrypted traffic to SaaS usage and API calls (with AI agents proliferating within enterprises), can persist with convergence. Converging multiple streams of telemetry without correlation can lead to data overload. Cross-domain or end-to-end root-cause analysis is just as hard because network/app/security data sit in separate systems — today’s vendor-specific AIOps rarely span multi-vendor environments.

And as we move toward a single policy engine governing every domain, the stakes increase, with unnecessary and undesired blind spots both within and across multiple networking and security domains.

Old Visibility Platforms Need a Refresh in a Converged World

I’ve previously written about the importance of modern traffic intelligence for today’s networks. Simplistic port/protocol-based classification is obsolete, and the lack of encrypted traffic intelligence is unacceptable now that an increasing amount of web traffic rides over HTTPS/QUIC. Similarly, over-reliance on decryption for inspection (unless it’s necessary for a full L7 content-aware proxy) is impractical and legally fraught. AvidThink’s same survey had 59% of respondents rank compliance among their top convergence challenges — GDPR, HIPAA, PCI-DSS, and the looming data sovereignty and privacy cloud that hangs over all enterprise CIOs mean much care is needed in decrypting potentially sensitive traffic.

What we do need, to support the dual axes of convergence (security + networking, multiple network domains), is continued investment in traffic visibility and observability capabilities to facilitate the integration of security into all network connectivity, and to enable correlation and tracking of traffic from end-to-end across multiple domains, over a mix of underlying transports (wireless, wired, non-terrestrial, fiber, fixed-wireless access).

Next-generation DPI and encrypted traffic intelligence remain critical amid the ongoing convergence that underpins today’s and tomorrow’s enterprise networks. It remains important as we wade into an era in which humans and agents generate traffic across AI-powered inference workloads that increasingly permeate all enterprise applications and software development processes. Likewise, traffic intelligence becomes particularly important with new AI-enabled attack capabilities including tunneling, domain fronting, anonymizers, and new attack surfaces opened up by a new class of cyber-capable models — such as Anthropic’s Claude Mythos, which was built for vulnerability discovery but carries clear offensive-misuse risk, and Anthropic itself expects “Mythos-class” rivals to follow within 6–18 months.

Traffic Intelligence Helps Eliminate Blind Spots in Converged Future Networks

Convergence is irreversible; we are seeing it first in mid-market, and larger enterprises with much more complex networks will follow in future refresh cycles. Small and medium enterprises will benefit from emerging all-in-one or managed solutions.

Blind spots, however, are unnecessary and avoidable. If we continue to invest in traffic visibility and demand improved visibility and correlation, even as enterprise networks converge along two axes, we can realize the potential benefits of OPEX savings and end-to-end management that convergence is supposed to bring.

For enterprises, we suggest making traffic visibility an explicit selection criterion. Networking leaders and CIOs should ask every vendor pitching convergence how they create cross-domain, end-to-end visibility and tracking. Likewise, they should ask how this improved visibility can help track ongoing developments with AI services, AI-generated traffic, and new AI-enabled attack vectors.

For vendors and carriers, treat embedded DPI as the enabling technology that can shine a bright light on traffic patterns (AI and non-AI, benign and malicious) even as your enterprise customers seek converged network management and unified policies. We recommend building on a proven visibility layer rather than reinventing one.

As enterprises pursue two axes of convergence, the goal is to help them collectively uncover a single source of truth.

Discover how Enea’s next-generation DPI and encrypted traffic intelligence give vendors and operators cross-domain visibility — including into encrypted and AI-generated traffic — across converged enterprise networks: https://www.enea.com/solutions/deep-packet-inspection-traffic-intelligence/

(1) North American medium size enterprises, n=120, survey gathered Q4 2025