A recent forecast paints a dramatic picture of the future, suggesting that global Wide Area Network (WAN) traffic could skyrocket by as much as 700% over the next decade, a surge primarily propelled by the insatiable data demands of Artificial Intelligence. This projection, put forth by industry heavyweight Nokia, envisions a world where network infrastructures are pushed to their limits by converging forces of consumer video, industrial automation, and exponential AI growth. However, this vision of a data deluge is being met with significant skepticism from seasoned industry analysts. They argue that the forecast, while ambitious, lacks a crucial connection to the tangible, real-world applications and current enterprise behaviors that would be necessary to fuel such an explosion, creating a compelling debate between a vendor’s forward-looking model and the grounded realities of the technology landscape.
A Vision of Unprecedented Data Demand
Nokia’s Global Network Traffic Report outlines a future where data volumes strain existing infrastructure, modeling its predictions across conservative, moderate, and aggressive scenarios that correspond to compound annual growth rates of 13%, 16%, and 22%, respectively. The forecast meticulously breaks down traffic into three primary domains, with the consumer sector, which includes mobile and fixed wireless access, projected to remain the single largest contributor. Within this domain, video content is expected to account for roughly 60% of all traffic, but with a significant evolution in consumption habits. The model anticipates a shift away from passive streaming toward more immersive, interactive, and “uplink-heavy” experiences, suggesting that users will become more active participants in media rather than just viewers. Concurrently, the report predicts a steady and substantial rise in traffic from the enterprise and industrial sectors, driven by the increasing digitalization of core operations. This vision includes highly connected factories where robots coordinate actions across different physical sites and field teams utilize Augmented Reality (AR) instead of traditional manuals.
Positioned as the undeniable “primary growth engine” for the entire forecast period, Artificial Intelligence is at the core of the predicted traffic explosion. The model projects that AI-related traffic will experience a staggering 23% compound annual growth rate, ultimately composing 30% of all global WAN traffic by 2034. This figure is all-encompassing, covering everything from consumer-facing assistants to sophisticated enterprise copilots and industrial automation systems. A key driver of this growth is the emergence of “agentic AI,” a concept where autonomous AI systems trigger other AI systems, generating a massive and continuous stream of machine-to-machine (M2M) traffic that operates in the background. Nokia emphasizes that this traffic will not be confined within data centers but will surge between them, creating new and demanding lateral data flows across edge, metro, core, and cloud infrastructures. For enterprise and industrial AI specifically, the forecast is even more aggressive, predicting a 48% CAGR and a fundamental shift from being primarily uplink-heavy (data collection) to becoming downlink-dominant as systems begin delivering immersive AI experiences back to users and machines.
A Grounded Counter-Narrative from Industry Experts
In stark contrast to these bold projections, independent industry analysts express profound doubts, primarily centered on the report’s failure to specify the applications that would necessitate such a massive increase in bandwidth. Tom Nolle, founder and principal analyst of Andover Intel, leads this critique by asserting that a traffic forecast is meaningless without a clear and verifiable link to the applications generating the data. He argues that when his own assumptions about future applications, which are based on extensive consultations with enterprises, are applied, Nokia’s numbers simply do not validate. This disconnect highlights a fundamental flaw in forecasting methodology: without grounding predictions in tangible use cases, they become abstract figures rather than actionable intelligence. This sentiment is further reinforced by the current priorities and concerns observed within the business world. Nolle, who engages with approximately 600 enterprises annually, reports a complete absence of concern among them that WAN bandwidth could become a bottleneck to achieving their strategic objectives. The notion that a CIO would readily accept a 30% surge in VPN traffic solely due to AI is dismissed as disconnected from financial realities.
The skepticism extends to the futuristic industrial scenarios painted in the forecast. Nolle challenges the premise of a “huge pent-up demand for cross-site coordination of IoT,” contending that if such a demand existed, there would already be widespread evidence of applications pushing the limits of current network capacity—a phenomenon he does not observe. Furthermore, the financial impracticality of rapid industrial transformation is a major hurdle. With the average depreciation timeline for industrial equipment exceeding 15 years, a swift replacement of legacy machinery with new, hyper-connected systems would force companies to absorb a significant financial loss, an action that would be severely punished by financial markets. William Webb, CEO of the consulting firm Commcisive, reinforces this view by pointing out the niche status of technologies like Augmented and Extended Reality. While some companies use them for specialized purposes like safety training, he notes that the “volume of devices is so small that the data generated is really insignificant globally.” He also finds scenarios like robots coordinating in real-time between different factories highly improbable, questioning the business case for such tight, millisecond-level coupling between geographically separate facilities.
Reconciling Forecasts with Reality
The central finding that emerged from this analysis was the significant chasm between a vendor’s market-shaping forecast and the on-the-ground reality reported by industry analysts. While experts acknowledged that growth was inevitable, the consensus was that Nokia’s projections appeared inflated and disconnected from the practical and financial constraints that enterprises face daily. A crucial piece of context was introduced by considering the source of the report. As a premier supplier of telecommunications equipment, Nokia was understood to have a vested interest in promoting a narrative of explosive growth in WAN usage. This commercial motivation suggested that their report might function not only as a forecast but also as a strategic tool designed to encourage investment in the very network infrastructure they sell. Webb lamented the absence of a truly authoritative, independent source for worldwide traffic forecasts, implying that reports from parties with a commercial stake should be approached with critical analysis. It was ultimately suggested that while AI and IoT would undoubtedly contribute to WAN traffic growth, the true catalyst remained the tangible, widespread adoption of applications, the pace and scale of which were still highly debatable.
