Steven Vondran, president and CEO of American Tower, said that the current phase of network deployment is increasingly focused on capacity rather than coverage
In sum – what to know:
5G densification accelerates – American Tower says carriers are shifting toward more colocation as they add network capacity, with densification expected to accelerate over time.
Spectrum fuels towers – The company expects additional spectrum to drive equipment deployments, while mobile data growth is increasing the need for network capacity.
AI expands demand – CoreSite is benefiting from AI inferencing today, while rising AI traffic could increase mobile-network investment and eventually require network re-architecture.
American Tower Corporation sees several reinforcing catalysts for its wireless infrastructure business, with 5G densification providing a near-term source of demand while additional spectrum, 6G, and AI could support investment over a longer period.
Speaking at the Goldman Sachs Communacopia + Technology Conference, Steven Vondran, president and chief executive officer of American Tower, said the company is seeing four catalysts for building in the business: densification, 6G, spectrum availability, and AI. The company’s first strategic priority for 2026 is capturing organic growth from 5G densification as carriers augment their networks.
The current phase of network deployment is increasingly focused on capacity rather than coverage, Vondran said. Carriers can add capacity through technology improvements and spectrum, but a significant portion comes through densification.
“So we’re already seeing that. We’ve seen a shift in the mix of our new business, a little bit less on the amendment side, more on the colocation side, but all underpinning a steady level of investment by the carriers. So the first catalyst is happening now, and we expect that to continue to accelerate over time,” he said.
Spectrum represents another potential multi-year catalyst. Vondran said 800MHz of spectrum has been earmarked to come to market over several years and argued that the historical relationship between spectrum and network equipment should continue.
“If you think about mobile data growth, it’s growing double digit or better. Network capacity needs to double by the end of the decade. Carriers will get some of that through technology upgrades. The rest is going to come from spectrum and site densification,” he added.
6G represents a longer-term catalyst. Vondran said standards are expected in 2029, with commercial deployments probably taking place around 2030 or 2031, while some activity could occur earlier through proofs of concept and early-stage networks.
AI is providing another source of demand, with the impact already visible in American Tower’s CoreSite business. Vondran emphasized that CoreSite is an interconnection hub rather than simply a data center company, with a customer mix spanning clouds, networks, and enterprises. AI and inferencing have increased the importance of multi-cloud and multi-inferencing environments, where enterprises connect directly to multiple tools.
American Tower has increased CoreSite capacity by about 1.5 times since acquiring the business and continues to invest in additional capacity. Vondran said demand is currently greater than what CoreSite can service and that interconnection revenue is also experiencing record growth.
AI is also expected to affect American Tower’s mobile-network business, although Vondran said AI traffic remains a small part of mobile traffic today. He expects that to change as AI usage expands and changes how people use mobile devices.
“I think it’s going to be a huge catalyst for us over the next decade. But while it’s a small piece of the pie, it’s a rapidly growing piece of the pie, and the uplink required by AI is more than what the networks are architected for today. So it could mean that there’s a network re-architecture that has to be done over time for that,” the executive said.
Facts Only
* Steven Vondran, president and CEO of American Tower, stated the current phase of network deployment focuses on capacity rather than coverage.
* 5G densification is accelerating, with carriers shifting toward more colocation as they add network capacity.
* Additional spectrum is expected to drive equipment deployments.
* Mobile data growth is increasing the need for network capacity.
* The company sees four catalysts for its wireless infrastructure business: densification, 6G, spectrum availability, and AI.
* The first strategic priority for 2026 is capturing organic growth from 5G densification.
* 800MHz of spectrum has been earmarked for market release over several years.
* Network capacity needs to double by the end of the decade due to mobile data growth, with capacity coming from technology upgrades, spectrum, and site densification.
* AI is currently benefiting CoreSite through AI inferencing.
* Demand for interconnection revenue at CoreSite is experiencing record growth.
Executive Summary
The current phase of network deployment is shifting focus from coverage to capacity, with 5G densification being a primary driver for this change, as carriers augment networks through technology and spectrum. This shift results in a movement toward colocation as part of capacity additions. The company anticipates that increased spectrum availability will drive equipment deployments, alongside the general growth in mobile data fueling the need for greater network capacity. Furthermore, the expansion of Artificial Intelligence (AI) is creating new demand streams, impacting both the interconnection hub business via CoreSite and potentially driving future investment in mobile networks through increased uplink requirements.
American Tower identifies four main catalysts for its wireless infrastructure business: densification, 6G development, spectrum availability, and AI. The immediate strategic priority for 2026 is capturing organic growth from 5G densification as carriers expand their networks. While AI's immediate impact on mobile traffic is small, the long-term expectation is that AI will necessitate network re-architecture over the next decade due to increased uplink demands. CoreSite is experiencing record growth in interconnection revenue, driven by demand for multi-cloud and multi-inferencing environments related to AI.
Full Take
The narrative centers on an inevitable evolution of network economics driven by escalating demand, specifically the need to move beyond simple coverage metrics to address capacity and specialized processing needs. The emphasis on 5G densification as the near-term catalyst frames infrastructure investment not just as a response to service demands but as a structural necessity dictated by technological layering. A key tension emerges between near-term operational drivers (like densification) and long-term disruptive forces (6G and AI).
The framing of AI is particularly telling; it moves from a current benefit within existing hubs (CoreSite) to a future driver for fundamental network re-architecture. This suggests that the investment thesis must account for not just physical capacity additions but also the necessary architectural shifts required to support emerging data flows, implying that hardware and site placement strategies will become increasingly dependent on sophisticated software integration over time. The discussion of spectrum links historical infrastructure needs directly to future capacity requirements, suggesting that resource allocation decisions are tightly coupled with anticipated technological rollouts rather than purely incremental deployment schedules.
What assumptions are being made about the pace of change? The text posits a linear progression where capacity must double by the end of the decade, driven by existing growth rates, and that AI will eventually mandate architectural changes. A deeper inquiry lies in whether this reliance on sequential drivers (5G $\rightarrow$ Spectrum $\rightarrow$ 6G $\rightarrow$ AI) correctly models the potential for parallel or emergent shifts in demand structures that might decouple these timelines. What are the non-linear interactions between spectral allocation, physical densification, and abstract computational demands?
Sentinel — Human
The text appears to be a faithful summary or transcription of an executive presentation, exhibiting characteristics consistent with human-generated analysis rather than pure generative synthesis.
