When AI Breaks the Enterprise Network — VERSA - October 28, 2025

By Staff Reports - October 28th, 2025

Artificial intelligence is rewriting the rules of enterprise connectivity. What began as a wave of automation and analytics has become a test of structural integrity: Can the network handle the complexity and velocity of AI-driven operations without breaking under its own weight?

The answer, says Kevin Sheu, vice president of product strategy at Versa, lies in rethinking how enterprises organize around performance, protection, and intelligence. In a recent discussion with BizTechReports, Sheu explained why Secure Access Service Edge (SASE) has evolved from an architecture choice into a strategic mandate—one that determines how well organizations can balance innovation with resilience.

In this Executive Q&A, Sheu explores how SASE changes the strategic landscape for enterprise technology leaders, the operational convergence reshaping IT and OT environments, the economics of risk management, and the technology development path required to sustain innovation in an AI-driven world.

Full Interview w/ Kevin Sheu, Vice President of Product Strategy, Versa

NOTE: This interview has been edited for clarity and length. The Q&A has been organized into four sections — Strategic Assessments, Operational Imperatives, Financial Implications, and Technology Development — to highlight the most pressing issues for today’s executives.

STRATEGIC ASSESSMENTS

 BTR: AI seems to be forcing a fundamental rethink of enterprise infrastructure. What’s changed about the network’s role in enabling digital transformation?

Sheu: Artificial intelligence has completely altered how we have to think about networks. They’re no longer utilities; they’re decision-making environments. Security, networking, and AI can’t operate as separate disciplines anymore. Your network has to support your security, your security has to protect the network, and both have to handle the traffic and data generated by AI.

What’s happening is that the network has become the control plane for innovation. It governs how fast you can deploy new digital services, how resilient you can be under pressure, and how well you can anticipate and mitigate risk. The traditional perimeter is gone. Now, the connection between performance and protection is the foundation of competitiveness.

BTR: What role does SASE play in this new reality?

Sheu: SASE brings the disciplines of networking and security together in one framework. It unifies software-defined wide-area networking (SD-WAN) with security service edge (SSE) functions—zero-trust network access, secure web gateways, data-loss prevention, and cloud access security brokers.

What used to be philosophies like zero trust and convergence are now operating principles, and that’s a big shift. When performance and protection rise together, enterprises can innovate faster without amplifying exposure. Unified connectivity and security mean AI-driven workloads can scale confidently and securely across multi-cloud environments.

BTR: AI has also changed the threat landscape. How do you see that dynamic evolving?

Sheu: It’s a double-edged sword. AI presents opportunities and challenges—and both are growing fast. Attackers are using it to become more sophisticated, while defenders are using it for predictive insight and rapid detection.

Because AI processes data volumes that the human mind—or even a patchwork of tools—can’t handle, it can close the gap between what organizations see and what’s actually happening. That’s where convergence matters most. The faster your systems can correlate and respond, the stronger your defensive posture becomes.

By removing the friction between speed and safety, SASE allows enterprises to innovate faster without amplifying exposure. Unified connectivity and security mean that new digital services, AI workloads, and remote operations can be launched with greater confidence and fewer dependencies. This acceleration, however, introduces its own demands: governance, culture, and collaboration must evolve just as quickly to sustain it.

OPERATIONAL IMPERATIVES

 BTR: What does convergence look like inside the enterprise?

Sheu: In the past, you had networking, security, and infrastructure teams each operating in their own lanes. That model no longer works. Networking, security, and AI operations can’t function as separate priorities anymore. Each depends on the others for performance, protection, and insight—if one fails, the rest fall behind.

The implication is clear: Without tighter alignment, enterprises risk building AI capabilities on unstable ground. Security gaps multiply, performance suffers, and innovation slows under the weight of disconnected systems.

We’re also seeing this convergence extend into operational technology (OT) environments—manufacturing, logistics, utilities. These are increasingly data-driven operations where AI is being deployed at the edge for predictive maintenance, energy management, and safety analytics. Historically, OT has been isolated from IT, but that separation is fading.

BTR: What challenges does that create for leaders managing both IT and OT domains?

 Sheu: First, you have to change how you think about visibility. Threats now move fluidly between on-premises systems, cloud platforms, OT networks, and remote users. You can’t protect one domain and ignore the others.

Second, you need a crawl, walk, run approach to automation and AI integration. Not every organization is ready to hand control over to autonomous systems. Start by letting AI assist—what we call “man in the loop.” Let human operators use AI to surface insights or recommendations. As you gain confidence, you can allow automation to execute more functions directly.

BTR: How important is cultural alignment to making this work?

Sheu: It’s essential. Technology convergence demands organizational convergence. If network and security teams aren’t aligned, your AI investments will underperform. This is where leadership comes in—creating a culture of collaboration and shared accountability. 

SASE gives you the platform, but people and process give you the momentum. Every company needs to define what convergence means for them, from workflows to metrics.

FINANCIAL IMPLICATIONS 

BTR: Beyond security and performance, there’s an economic argument for SASE. How does this model change the cost structure of risk management?

Sheu: The economics are being rewritten. The old model—multiple tools, multiple policies, multiple points of failure—is expensive and inefficient. SASE simplifies that by unifying control under one policy fabric.

It introduces a new economic calculus: fewer platforms to manage, less duplication of effort, and lower operational risk. But the real ROI comes from resilience and agility. Downtime, data loss, and compliance failures cost far more than software licenses.

At Versa, we think of this as a two-way relationship between AI and SASE—AI for SASE, and SASE for AI. On one side, you embed AI into security and networking to enhance detection and response. On the other, you modernize the network to handle the massive data flows that AI generates. Together, those investments support a self-reinforcing cycle of performance and protection.

BTR: What kind of ROI should business leaders expect to see?

Sheu: It depends on where they start. But we typically see measurable impact in four areas:

•  Automation and AI copilots reduce manual workload and improve incident response times.

• Intelligent traffic engineering keeps data close to inference and training sources, minimizing latency.

• Unified policies limit the blast radius of security incidents.

• Cloud-delivered architecture eliminates the need for constant hardware refreshes.

But I would stress this: SASE is not just about cost savings. It’s about enabling innovation safely. If you can’t secure your AI, your innovation will stall.

TECHNOLOGY DEVELOPMENT

BTR: What’s the technology roadmap for SASE in the context of AI evolution?

Sheu: The most exciting development is what we call the AI fabric—the interconnection of AI workloads, data flows, and security controls across every environment. Organizations are realizing that they can’t manage AI in isolation. The network itself becomes the intelligent backbone that supports inferencing, governance, and data movement.

At Versa, we think of AI in three domains: AI for security, AI for networking, and AI for operations. Each is advancing rapidly.

• AI for security improves detection and prediction—identifying threats that human analysts might miss.

• AI for networking optimizes traffic flows and allocates resources dynamically based on real-time conditions.

• AI for operations enhances visibility, predictive maintenance, and human-AI collaboration through copilots.

 At the same time, SASE for AI focuses on the reverse relationship—how networking and security architectures are redesigned to enable AI workloads. As inference moves closer to the edge, performance and governance need to scale with it.

BTR: How can enterprises prepare for that shift?

Sheu: Think of it as building infrastructure that learns. Start by assessing where your AI workloads reside and how data moves between endpoints. Then look at your network architecture: Can it prioritize and secure AI traffic without bottlenecks?

AI traffic is different—it’s more dynamic, more data-heavy, and often more sensitive. So, we’re helping organizations design for intelligent resilience—networks that not only perform but adapt.

Ultimately, the companies that thrive will be those that approach AI and SASE as one problem set, not two.

BizTechReports Perspective:

 The convergence of networking, security, and artificial intelligence marks a pivotal moment in enterprise architecture. The emergence of SASE as a unifying framework gives organizations the flexibility to move fast and the confidence to innovate safely. 

Kevin Sheu’s insights illustrate that the path forward is not about replacing existing systems overnight but building adaptive foundations that can evolve with AI’s accelerating demands. The result is a model where resilience and agility reinforce each other—a new standard for operational intelligence.

As governance frameworks mature and automation becomes more pervasive, enterprises that unify these disciplines today will be best positioned to lead tomorrow’s intelligent economy.