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 shift from experimental generative AI to widespread deployment of AI agents is fundamentally altering how networks are designed, secured, and managed.

A Cisco and Foundry survey of 3,472 IT and networking leaders across 15 countries found AI is already changing traffic patterns across campus and branch environments and exposing capacity, security, and visibility gaps that many organizations aren’t prepared to address. The findings highlight a critical blind spot in AI readiness: the network edge where employees, devices, and applications connect.

“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 data underscores that enterprises may need to expand AI readiness planning beyond data centers and cloud environments and pay more attention to the networks connecting employees, applications, and devices. This issue will become more significant as enterprise organizations move beyond generative AI pilots and begin deploying AI agents that communicate continuously with other systems and applications.

Key Survey Findings

The Cisco survey revealed several critical statistics that paint a clear picture of the growing strain on campus and branch networks:

• Organizations reported a 34% increase in AI-related campus and branch network traffic over the past 12 months.
• Traffic is projected to climb 209% over the next three years, with companies broadly deploying AI expecting total network traffic to triple.
• 73% already face, or expect to face, campus and branch network capacity constraints within the next two years.
• 67% said AI workloads are increasing east-west traffic between internal systems and applications.
• 80% said AI has expanded their attack surface.
• 61% said they are delaying additional AI deployments until they gain more confidence in their security posture.
• 85% expect moderate or significant growth in AI agent deployments over the next two years.

These numbers indicate that the traditional approach of designing networks for predictable, north-south traffic patterns—such as SaaS and CRM access—is no longer sufficient. AI agents and generative AI applications introduce bursty, unpredictable flows that can overwhelm legacy architectures.

Changing Traffic Patterns and Observability Gaps

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?”

The shift to east-west traffic—data moving between servers and applications within the same network—requires a different network architecture. Traditional hub-and-spoke designs optimized for internet-bound traffic fail to handle the high-volume, low-latency communication needed for AI agents to collaborate effectively. Network engineers must now consider leaf-spine topologies, higher bandwidth links, and microsegmentation to support this new traffic pattern.

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 includes investments in faster switches, wireless access points capable of handling dense AI client traffic, and software-defined networking overlays that can dynamically route traffic based on workload priorities.

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.”

Lack of visibility into AI traffic makes capacity planning, security monitoring, and compliance enforcement nearly impossible. Enterprises need to deploy network detection and response tools, as well as AI-powered observability platforms that can automatically classify and prioritize AI workloads. Without such tools, network teams are flying blind as AI adoption accelerates.

Security Implications of AI Traffic

Security is also 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 expanded attack surface—reported by 80% of survey respondents—stems from several factors. AI agents often require direct access to internal databases, APIs, and other sensitive resources, creating new vectors for data exfiltration. Additionally, employees using unsanctioned generative AI tools can inadvertently leak proprietary information or introduce vulnerabilities through third-party plugins. Network teams must implement zero-trust architectures, secure access service edge (SASE) frameworks, and AI-specific security policies that restrict what data AI agents can access and how they communicate.

Furthermore, 61% of organizations are delaying additional AI deployments until they gain more confidence in their security posture. This hesitation could slow competitive advantage, but it reflects a prudent approach. The complexity of securing AI workflows—especially when agents interact autonomously—requires new security controls such as AI firewalls, runtime application self-protection, and continuous monitoring of model behavior. Enterprises that fail to address these security gaps risk data breaches, compliance violations, and reputational damage.

The Path Forward: Network Modernization

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. 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.

Network modernization efforts must address several key areas. First, capacity: upgrading to Wi-Fi 7, 25GbE or 100GbE edge switches, and higher-bandwidth WAN connections to handle the predicted 209% traffic increase. Second, architecture: moving from traditional three-tier designs to software-defined networking that supports dynamic traffic steering and microsegmentation for AI workloads. Third, observability: deploying network performance monitoring and AIOps tools that provide real-time visibility into AI traffic patterns and application performance. Fourth, security: implementing SASE, zero-trust network access, and AI-specific security policies to protect the expanded attack surface.

Organizations that invest now in network readiness will be better positioned to capitalize on AI’s full potential. Those that delay risk falling behind as AI becomes a core business driver. Patel said in the statement: “Eventually there will be only two kinds of companies: those that are AI companies, and those that are irrelevant.”

The survey also underscores the importance of workforce skills. Network engineers must upskill to understand AI workloads, traffic patterns, and security requirements. IT leaders should invest in training, certification programs, and cross-functional collaboration between network, security, and AI teams. Without the right expertise, even the most advanced network infrastructure will fail to deliver the performance and reliability that AI applications demand.

For many organizations, the journey toward AI-ready networks has just begun. The Cisco research serves as a wake-up call that the network is no longer a passive utility but a strategic asset that must evolve in lockstep with AI innovation. The next two years will separate companies that successfully integrate AI into their operations from those that struggle with capacity, security, and visibility issues.

Source: Network World News