While enterprise IT leaders have spent the past two years focusing AI infrastructure discussions on GPUs, cloud platforms, and data centers, new Cisco research suggests that enterprise networks may not be ready for the next phase of AI adoption. The findings, drawn from a Cisco and Foundry survey of 3,472 IT and networking leaders across 15 countries, paint a sobering picture: AI is already reshaping traffic patterns in campus and branch environments, exposing critical capacity, security, and visibility gaps that many organizations are not equipped to address.
“We have entered a networking supercycle, because the network is so central to all the AI infrastructure the world is building now,” said Jeetu Patel, Cisco president and chief product officer, in a statement. The survey results underscore that enterprises may need to broaden their AI readiness planning far beyond data centers and cloud environments. The networks connecting employees, applications, and devices are becoming just as crucial to AI success, especially as organizations move beyond generative AI pilots and begin deploying AI agents that communicate continuously with other systems and applications.
The survey found that organizations reported a 34% increase in AI-related campus and branch network traffic over the past 12 months. That figure is only the beginning: traffic is projected to climb 209% over the next three years. Companies that are broadly deploying AI expect total network traffic to triple. Despite this explosive growth, 73% of respondents already face, or expect to face, campus and branch network capacity constraints within the next two years. Moreover, 67% said AI workloads are increasing east-west traffic between internal systems and applications—a traffic pattern that traditional hub-and-spoke networks are not designed to handle efficiently.
The security implications are equally stark. A full 80% of IT leaders said AI has expanded their attack surface, and 61% reported they are delaying additional AI deployments until they gain more confidence in their security posture. These figures highlight a growing tension between the desire to adopt AI at scale and the practical challenges of securing an increasingly complex network environment. The report also noted that 85% of respondents expect moderate or significant growth in AI agent deployments over the next two years, adding further pressure to already strained network architectures.
Changing traffic patterns inside enterprise environments are causing additional pressure for network teams. “Usually, networks are designed for consistent traffic, like SaaS and CRM traffic, and there aren’t a lot of unpredictable traffic patterns,” said the head of AI strategy for global IT and network engineering operations at a large U.S. technology company who participated in the research. “Suddenly, three AI agents are trying to talk to each other and solve a problem. That is going to be a big thing … how do we support increased east-west traffic?” This question resonates across industries as AI agents—autonomous software entities that perform tasks, interact with other systems, and make decisions—become more prevalent. Traditional network designs, which often rely on centralized data centers and limited bandwidth for lateral communication, are being stretched to their limits.
Cisco defined aggressive AI adopters as organizations with broad generative AI deployments across the enterprise, but only 30% of those organizations said they are fully prepared to support projected AI growth across their networks. As a result, 93% of IT decision makers said they are accelerating network modernization efforts. This modernization push is not just about adding more bandwidth; it involves rethinking network architecture, adopting intent-based networking, implementing zero-trust security models, and investing in AI-driven network management tools. The need for real-time visibility into traffic patterns and application performance has never been greater.
The report also highlighted an observability challenge that could complicate future deployments. As employees and business units increasingly experiment with AI tools, IT organizations may not know what is actually running on their networks. “Right now, we don’t even know what the AI-driven demand is,” the AI strategy executive said. “Observability is a huge gap. There is experimentation going on all over the place, and there is no way for us to really identify if somebody is deploying some kind of service on our network, whether it is a genAI solution or an agentic solution.” This lack of visibility creates risks not only for capacity planning but also for security and compliance. Without a clear picture of which AI workloads are running, where they are hosted, and how they interact, IT teams cannot effectively manage performance, troubleshoot issues, or enforce governance policies.
Security is emerging as a barrier to AI expansion as organizations struggle to govern rapidly growing numbers of AI tools and workloads. “The issue from a security standpoint is that it’s hard to create the guardrails for every possible AI tool that your organization must use,” said the vice president of infrastructure, network, and end-user services at a U.S. retail enterprise interviewed for the report. The proliferation of AI-powered applications—from chatbots to code assistants to data analytics platforms—means that traditional perimeter-based security models are no longer adequate. Network segmentation, micro-segmentation, and continuous monitoring are becoming essential to prevent lateral movement by attackers who might exploit AI workloads. Additionally, AI systems themselves can be vectors for attacks, such as adversarial machine learning or data poisoning, requiring specialized security measures.
The AI readiness conversation has often centered on data centers, but AI applications operate where employees work, devices connect, and business processes run. That means campus and branch environments may become just as important to AI success as the infrastructure supporting AI models. For example, a manufacturing plant using AI for predictive maintenance needs a network that can handle real-time sensor data from thousands of IoT devices. A retail chain deploying AI-powered inventory management must ensure that its branch networks can support the sudden spikes in data traffic when AI agents analyze stock levels and trigger replenishment orders. Similarly, remote workers using AI collaboration tools require reliable, low-latency connections to cloud-based AI services.
The Cisco research shows that AI infrastructure planning can no longer focus only on back-end systems if enterprises expect to scale AI deployments over the next several years. Patel expressed this urgency in his statement: “Eventually there will be only two kinds of companies: those that are AI companies, and those that are irrelevant.” The survey serves as a wake-up call for IT leaders to start evaluating their campus and branch networks now, before AI traffic overwhelms them. Investments in network modernization—such as upgrading to Wi-Fi 6/7, deploying software-defined wide area networking (SD-WAN), implementing AIOps platforms for proactive monitoring, and adopting SASE frameworks for secure access—are likely to become prerequisites for successful AI adoption.
In addition to the technical challenges, the survey points to a cultural shift within IT organizations. Network teams that traditionally focused on maintaining uptime and reacting to incidents are now being asked to become strategic partners in AI initiatives. This requires new skills in data science, cloud networking, and security analytics. Many organizations are struggling to find talent with these hybrid skills, further complicating their AI journeys. The survey did not delve into the workforce implications in detail, but the pressure on IT teams is clear: they must balance day-to-day operations with the need to redesign networks for an AI-driven future.
The findings also have implications for vendors and service providers. Cisco, as the survey sponsor, is positioning itself as a leader in AI-ready networking, but the results suggest that no single vendor can address all the gaps. Enterprises will likely need a combination of hardware upgrades, software tools, and managed services to bridge the visibility, security, and capacity deficits. The emergence of AI agents—which require low-latency, high-bandwidth connections to multiple data sources—will only intensify these requirements. Organizations that fail to adapt may find themselves unable to deploy AI at scale, putting them at a competitive disadvantage.
As AI continues to transform enterprise operations, the network remains the backbone of digital transformation. The Cisco survey provides a timely reality check: while much of the AI conversation has focused on the compute and storage side, the network that connects everything is often an afterthought. For enterprises to fully realize the benefits of AI—automation, insights, efficiency—they must ensure their campus and branch networks are not a bottleneck. The next phase of AI adoption will test whether organizations can align their network strategies with their AI ambitions, and the results of that test will likely determine which companies thrive and which are left behind.
Source: Network World News