AI is fundamentally changing the way enterprise and service provider networks operate, and network operators must adapt quickly to support AI workloads, a Cisco executive told a U.S. Senate subcommittee last week. Bob Everson, chief architect of provider mobility at Cisco, testified before the Subcommittee on Telecommunications and Media, part of the U.S. Senate Committee on Commerce, Science, and Transportation. His remarks focused on two core questions: how AI is reshaping networks, and how networks can leverage AI to become more intelligent and resilient.
The July 30 hearing, titled “Intelligent Networks: Powering Artificial Intelligence and Transforming Communications,” was convened to examine how the rapid adoption of AI has impacted network infrastructure. In her opening statement, Senator Deb Fischer (R-Neb.), chairman of the subcommittee, emphasized that widespread AI use has forced networks to evolve, requiring more capacity and more complex designs. She also noted that private companies have invested hundreds of billions of dollars in network deployment in recent years, and that various federal broadband programs have provided billions more to support targeted deployment and maintenance across the country.
Everson was joined by witnesses from U.S. Telecom, Vanderbilt University, and the Nebraska Public Service Commission. The panel heard a range of perspectives on how AI is transforming communications, but Everson’s testimony stood out for its detailed technical insights and data-driven observations about network traffic patterns.
AI's Growing Impact on Network Traffic
Everson began by describing how AI is changing not just the volume of network traffic, but its behavior. He noted that Cisco measured a fourfold increase in AI inference traffic over just eight months, a clear signal that AI workloads are becoming a dominant force in network design. Traditionally, networks have been optimized for content flowing downstream—think video streaming and web browsing. But AI is far more two-way and uplink-intensive. Prompts, context, sensor data, and agent activity all travel back toward AI models, and the resulting connections are active much longer than conventional web transactions.
“AI agents amplify these effects by operating at software speed,” Everson said. “In our testing, an agent generated 450 percent more traffic than a person performing the same task, and roughly 70 percent of that additional traffic was inference.” This means that as AI agents become more common in enterprise workflows, network planners must rethink their assumptions about capacity and latency.
Everson provided specific examples from real-world deployments. In campus and branch networks—like the one powering the Senate office building where the hearing took place—customers have already reported a 34% increase in traffic tied to AI workloads over the last 12 months, and they expect to see a 96% increase in the coming year. Half of enterprise customers report that AI demand is concentrated on their Wi-Fi networks, and 73% of organizations already face or expect to face campus and branch capacity limitations within the next 24 months.
These figures underscore the growing pressure on network infrastructure, especially as AI workloads become more distributed. Everson explained that while the large majority of AI to date has come from foundation models running on central infrastructure, enterprises are increasingly deploying small language models, open-source models, and specialized models—such as vision and voice models—which can be distributed throughout the network. Each of these characteristics, he said, highlights the value of the FCC’s forward-thinking decision in 2020 to authorize the full 6 GHz band for unlicensed Wi-Fi use.
Key Challenges: Infrastructure, Cost, and Security
In his prepared remarks, Everson identified several areas that are particularly impacted by AI. The first is infrastructure. AI is driving a shift toward edge computing, where data is processed closer to the source rather than in a centralized cloud. Service providers must consider “AI-native” traffic profiles for several reasons, including technical considerations, cost, and data sovereignty and security issues.
On the technical side, physical AI use cases such as robotics, autonomous vehicles, and industrial automation could require sub-millisecond decision-making. If an autonomous robot sends data to a central cloud and has to wait for a response, the round-trip latency could be too high for safe, real-time operation. Edge computing is therefore not just a convenience but a necessity for many AI applications.
Cost is another major factor. AI operations generate massive amounts of data. For example, high-definition video analytics for public safety can generate terabytes of data daily. Backhauling that data to a central cloud is prohibitively expensive and creates massive network congestion. Everson argued that processing data closer to the edge is not only more efficient but also more economical.
Data sovereignty and security are also critical concerns. Enterprises and governments are increasingly worried about moving sensitive information across the public internet to a third-party cloud provider. Many customers have security or regulatory concerns that prevent them from using centralized cloud infrastructure for certain AI workloads. By distributing computing resources to the edge, organizations can maintain better control over their data and comply with local regulations.
Potential Benefits of AI-Driven Networks
While AI workloads present challenges, Everson was quick to point out that networks can also leverage AI to boost performance and resiliency. He described the concept of AgenticOps, which allows the network to act as a self-healing system. Cisco’s AI-native tools, he said, enable the network to reroute traffic, adjust capacity, or reconfigure network nodes when the system detects performance degradation or an impending hardware failure. This dramatically increases uptime and reliability for mission-critical services.
AI is also helping to close the talent gap that many network operators face. Managing increasingly complex, software-defined networks requires specialized skills that are in short supply. AgenticOps allows operators to automate repetitive, low-value tasks—such as ticket resolution, configuration updates, and routine maintenance. These tools help lower the barrier to entry, allowing more junior analysts to ramp up quickly. By automating these tasks, Cisco’s AI-enabled platforms can free network engineers to focus on higher-level architectural strategy and innovation, and free cybersecurity analysts to dedicate more time to strategic threat hunting and detection engineering.
Another promising development is Integrated Sensing and Communication (ISAC), which combines wireless communications and radio-frequency sensing to “see” objects’ position and path using radio waves that reflect off them. Unlike optical sensors, ISAC can detect intrusion even in low-light conditions, through smoke, or around obstructions where traditional video analytics might fail. Everson said this technology has been prototyped and demonstrated already, and it holds great promise for autonomous systems and robotics, AI-driven smart facilities, and public safety.
Policy Recommendations for an AI-Native Future
Everson concluded his testimony by offering three suggestions for the committee to consider as they work to support AI-driven networks. First, he called on Congress to accelerate the U.S. AI-native stack. Cisco is investing across multiple dimensions of AI native networking, bringing new capabilities to 5G-Advanced today while building the foundation for 6G. One example of this commitment is AI-WIN—a collaboration among Cisco, NVIDIA, MITRE, Orion Development Company, Booz Allen, and T-Mobile—which brings AI, compute, and wireless together to create a secure, American-led path from 5G-Advanced to AI native 6G. “I encourage Congress to lean in on areas where the United States has a strategic leadership role, such as compute, core networking, and applications,” Everson stated.
Second, he urged policymakers to modernize permitting and infrastructure processes. As computing becomes more distributed, permitting must enable rapid and responsible deployment. As this Committee considers the future of the Universal Service Fund, it should account for the evolving costs of AI-ready networks so rural and urban communities can share in the benefits. The expansion of broadband infrastructure is critical to ensuring that the benefits of AI are widely distributed, and outdated permitting processes can slow down the deployment of new fiber, small cells, and other network upgrades.
Third, Everson stressed the importance of maintaining a balanced spectrum policy. The 800 megahertz of licensed spectrum recently made available by Congress is essential to high-capacity, high-uplink connectivity. The FCC’s authorization of the 6 GHz band for unlicensed use is equally important to meeting enterprise demand. He thanked the committee for its efforts to rebuild “a dependable pipeline of both,” which he described as foundational to American leadership in AI and networking.
The hearing highlighted the growing consensus among industry, academic, and regulatory experts that AI is not just another application running on the network—it is a fundamental driver of network architecture. As Everson put it, “AI is changing not only the volume of network traffic, but the behavior.” The shift to two-way, uplink-intensive communication, the explosion of data from edge devices, and the need for ultra-low latency are all forcing network operators to rethink everything from spectrum allocation to data center design.
With enterprise customers already reporting significant traffic increases linked to AI workloads, and with many more expecting to see capacity limitations within the next two years, the pressure is on for providers, policymakers, and enterprises to act. The solutions will require a combination of advanced technologies, forward-looking regulation, and clever use of AI itself to make networks more efficient and resilient. As the hearing made clear, the future of networking is intimately tied to the future of AI, and the decisions made today will shape the digital infrastructure of tomorrow.
Source: Network World News