While enterprise IT leaders have spent the past two years focusing AI infrastructure discussions on GPUs, cloud platforms, and data centers, new research from Cisco and Foundry suggests that enterprise networks may not be ready for the next phase of AI adoption. The survey of 3,472 IT and networking leaders across 15 countries found that AI is already changing traffic patterns across campus and branch environments, exposing capacity, security, and visibility gaps that many organizations are unprepared to address.
According to the report, organizations reported a 34% increase in AI-related campus and branch network traffic over the past 12 months. That figure is projected to climb 209% over the next three years, with companies that are broadly deploying AI expecting total network traffic to triple. The rapid growth is forcing network teams to reevaluate traditional design assumptions. Networks have historically been optimized for consistent north-south traffic patterns—data flowing between users and centralized applications—but AI workloads are increasingly generating east-west traffic between internal systems and applications. 67% of respondents said AI workloads are increasing this type of lateral traffic, which can overwhelm switches and routers designed for more predictable flows.
The capacity crunch is a pressing concern: 73% of organizations already face, or expect to face, campus and branch network capacity constraints within the next two years. Despite the urgency, only 30% of aggressive AI adopters—those with broad generative AI deployments—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 investment is not limited to data center upgrades; it encompasses edge switches, Wi-Fi 7 access points, and software-defined networking overlays that can dynamically adapt to changing traffic demands.
Security is emerging as another major barrier to AI expansion. 80% of respondents said AI has expanded their attack surface, and 61% said they are delaying additional AI deployments until they gain more confidence in their security posture. The challenge lies in the sheer variety and volume of AI tools entering the enterprise. Shadow IT and departmental experimentation with generative AI and agentic solutions make it difficult for security teams to create guardrails for every possible tool. One retail IT executive quoted in the report noted that it is hard to enforce policies when employees can spin up AI services without oversight. The survey also highlighted an observability challenge: many IT organizations do not know what is actually running on their networks. "Right now, we don't even know what the AI-driven demand is," said one technology company executive. "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." This lack of visibility complicates capacity planning, security monitoring, and cost management.
The rise of AI agents is expected to amplify these issues. 85% of respondents expect moderate or significant growth in AI agent deployments over the next two years. Unlike static applications, AI agents communicate continuously with other systems and applications, creating unpredictable bursts of east-west traffic. Traditional network architectures that rely on fixed topologies and manual configuration struggle to accommodate this dynamic behavior. Network teams are therefore exploring intent-based networking, AI-driven automation, and real-time traffic analytics to keep pace. The findings underscore that AI infrastructure planning can no longer focus only on back-end systems like GPU clusters and data centers. AI applications operate where employees work, devices connect, and business processes run—primarily in campus and branch environments. As Jeetu Patel, Cisco’s president and chief product officer, stated in the report: "We have entered a networking supercycle, because the network is so central to all the AI infrastructure the world is building now." He added that eventually there will be only two kinds of companies: those that are AI companies, and those that are irrelevant.
The Cisco research provides a wake-up call for enterprise IT leaders who have concentrated AI readiness efforts on the data center. The next wave of AI adoption will demand equally robust campus and branch networks, backed by enhanced security observability and the agility to handle unknown traffic patterns. Organizations that fail to modernize their local networks risk stalling AI deployment and falling behind competitors.
Source: Network World News