Will the U.S. Government and Hyperscalers Push the Mobile Industry to the Forefront of 6G?

2025 Brooklyn 6G Summit, Brooklyn, New York, Nov. 5-7, 2025 — More than 300 in-person attendees and 1,600 virtual attendees from academia, technology standards bodies, the public sector, industry analyst firms, network infrastructure and device vendors, communication service providers (CSPs), satellite network operators, semiconductor firms, hyperscalers and other stakeholders of the broader wireless technology ecosystem gathered at the Tandon School of Engineering at New York University in Brooklyn for a mix of presentations, fireside chats and panel discussions regarding the future of wireless networks, with a focus on 6G. There were more than 35 exhibits on display from graduate students at various U.S. universities as well as from a range of telecom ecosystem entities showcasing 6G-related innovations. A broad range of other 6G-related topics were discussed, including the role of AI and machine learning in networks, sustainability (especially related to energy), standards development, spectrum policy, public-private collaboration, and nonterrestrial networks’ role in the ecosystem. The 12th annual event was hosted by Nokia and NYU WIRELESS. This year’s event provided the usual cellular ecosystem updates, but one of the major themes was how the telecom industry should approach converting wireless innovation into value, especially economic value. Bell Labs’ 100th anniversary was also commemorated at the event.

TBR perspective

There is no doubt that the global economy is in the midst of an AI super cycle. What is doubtful, though, is whether the telecom industry will rise to the occasion to support (and monetize) opportunities that arise from AI. There are several fundamental challenges endemic to the telecom industry that could keep it from participating in the AI economy in a significant way, and TBR believes the AI economy could disintermediate at least a portion of the telecom industry.

The AI ecosystem currently operates in 18-month innovation cycles, and the telecom industry remains mired in its decade-long generational cycles; this dissonance will create friction and could be the biggest impediment to the convergence of the AI economy with the telecom industry. Will the rapidly evolving AI ecosystem wait for the telecom industry to support it, or will it be forced to move beyond telecom incumbents and institutions to avoid being held back?

One thing is certain: AI will fundamentally change how networks are utilized, necessitating a new network architecture. The telecom industry, which includes the cellular ecosystem, moves incredibly slowly, and this slowness is diametrically opposed to how the AI ecosystem (and other tech theme areas) moves. Something is going to break, and the real questions are what and when.

Impact and opportunities

A new network architecture is required for AI, as current networks will not suffice

One key aspect of AI workloads, especially those emanating from end-user devices, is that they are uplink-intensive, meaning they rely more heavily on uplink resources from the network than on downlink resources. This is a fundamental issue because macro, cellular-based networks are optimized for downlink capacity (typically a 10-1 downlink-uplink ratio from a resource-allocation perspective). CSPs will need to make significant investments in new network technologies and rethink how spectrum resources are utilized, to optimize networks for uplink.

AI traffic also tends to require lower latency than current networks and can support higher bursts of traffic than video and other media consumption. AI networks require uplink bandwidth, lower latency (compared to current networks) and the ability to handle higher bursts in traffic patterns at scale, and none of these requirements can be achieved just by increasing capacity. These requirements are the opposite of how networks are architected today — optimized for downlink, best-effort or good-enough latency, and optimized for more predictable traffic patterns — necessitating significant investment by CSPs. This will be a gradual transition, as there is no silver bullet to address this problem quickly. The best approach seems to be decoupling the downlink from the uplink for transmit power differentials, which would enable network resources to dynamically adapt to traffic demands in real time. Additionally, there is concern as to how willing CSPs will be to invest in uplink when ROI is uncertain.

6G will primarily be a software upgrade

Unlike prior Gs, 6G is unlikely to be a massive hardware refresh, but will instead build on top of existing 5G RAN and 5G core (the 5G SA [standalone] architecture) infrastructure as a software upgrade. This is the most likely outcome as CSPs are highly unlikely to have the appetite or the financial wherewithal to invest in another massive network refresh with unclear ROI.

There will be some hardware elements to invest in, such as processing capability that enhances programmability and maximizes AI support, or radios that support new frequencies not covered by existing RAN via software-defined radio technology, but hardware investments will be a fraction of what was spent in prior cellular generations. As a result, TBR expects a shallower capex curve during the 6G cycle, more closely resembling the 5G Advanced spend curve thus far.

Trump likes 6G

The Trump administration has designated several technologies and resources as strategic national priorities. Much as he did during his first administration, designating 5G as a national strategic priority, Trump views 6G as equally, if not more important. 6G will be leveraged for a broad range of defense and public safety, as well as societal use cases, which makes investment and success with the technology a public issue. Integrated sensing and communication (ISAC) is a 6G-related use case of particular interest to the U.S. government, for national security considerations. The government is heavily involved in 6G R&D-related endeavors, both directly and indirectly, through proxy programs such as the Defense Advanced Research Projects Agency and the U.S. National Science Foundation, as well as government agencies such as the U.S. Department of War.

TBR expects the Trump administration to apply similar approaches to seeding and bolstering U.S. innovation in 6G, much like it has in the production of rare earth metals, AI, nuclear power, and other key technologies. As part of this, it is reasonable to expect U.S. government equity investments and policy support. The recent Nokia deal with NVIDIA and the U.S. and Nokia agreement both align with this policy.

Cellular standards fracturing goes beyond geopolitics

Cellular standards are not only being regionalized due to geopolitical and nationalistic considerations but are also showing cracks within regional ecosystems. For example, non-standardized technologies are increasingly coexisting with 3rd Generation Partnership Project (3GPP)-based cellular technology. Standards in cloud data centers, optics, chip design and end-user devices are fragmenting, creating technological walled gardens.

For example, hyperscalers now have enough global scale to justify pursuing their own technology standards, and it is reasonable that these companies could, at some point, rival or eclipse traditional standards bodies, such as the 3GPP, in ecosystem influence and market power. Potential catalysts for a change such as this could include the need for AI-native networks (thereby reducing dependencies on CSPs for infrastructure investment and innovation road map alignment) and the need to support the rapidly evolving AR/VR market, which holds multitrillion-dollar revenue potential for hyperscalers.

Unlocking stranded spectrum assets: A prerequisite for 6G leadership

The mobile industry continues to beat the drum for more spectrum, but it should instead focus on fully utilizing the spectrum already allocated. TBR notes there are vast tranches of spectrum in the U.S. market that are broadly underutilized, either for technical or economic reasons. And challenges will only worsen as the industry aims to bring upper midband frequencies into the fray, which have greater propagation challenges and are less suited for macro coverage.

The U.S. needs to do a better job, guided by government institutions like the Federal Communications Commission, of utilizing CBRS, C-Band, 6GHz and mmWave bands, which are woefully underutilized today. For example, only a relatively small portion of midband spectrum has been deployed in the U.S. market to date, implying that well over half of it has not yet been put to use. (CSPs are either sitting on it or hoarding it.) Most 4G and 5G network traffic in the U.S. today runs over low bands such as 600MHz to 800MHz and the lower midband (1GHz to 2.6GHz, especially 2.5GHz), with C-Band increasing but nowhere near its full utilization potential. MmWave bands hold promise, but for economic and technical reasons, they were used only in very specific situations, mostly for LAN capacity.

Additionally, spectrum warehousing entities like EchoStar (which recently reluctantly agreed to sell a portion of its vast spectrum holdings to Starlink and AT&T) and private equities are still sitting on large tranches of unused spectrum. The government should redirect its efforts toward addressing these market dislocations rather than straining to bring new spectrum to a market that might not even be used (case in point, CBRS). This should be a mandatory government push because dislocations created by negative externalities of capitalism threaten to keep the U.S. behind China in key technological domains; specifically, corporations with a scarcity mindset are hoarding, but not necessarily using, the spectrum resources they have. The U.S. is already behind China across several key technologies (e.g., 5G, hypersonic missiles, electric batteries and vehicles, rare earth element processing); it is unacceptable for the U.S. to also fall behind on 6G.

TBR believes that the spectrum already allocated to CSPs will be deployed once the business case for its use is secured. Spectrum resources will come out of the shadows once ROI is clear. The government can help in this regard.

Data protectionism is stifling innovation and holding back the U.S.

A major limitation for academia and the broader mobile ecosystem is the lack of raw data to leverage for innovation. Data is viewed as a strategic and valuable asset, and the reality is that companies and governments do not want to share their data for various reasons. This limits academia, research institutions and other companies in what can be used for educational purposes as well as their ability to innovate by training their models and fine-tuning their technologies. Here again, China has an advantage over the U.S. because its authoritarian model can accelerate the pace of innovation in targeted areas.

Conclusion

The AI and cellular ecosystems are moving at different speeds rather than converging onto parallel tracks, which is a major issue and could lead to a breaking point for the technology industry. If hyperscalers feel like CSPs are holding them back, they can and will, as history validates, work around CSPs and take greater control over their own destinies.

This could include building their own standards for network infrastructure, much like they have for IT infrastructure and at the endpoint-device level and build out AI-optimized networks far greater in scope than they have been building (see TBR’s Hyperscaler Digital Ecosystem Market Landscape and Hyperscaler Capex Market Forecast reports for more information). Since it is unlikely the telecom ecosystem will fundamentally change, as it is not geared to do so, greater disintermediation and changing competitive dynamics are likely to occur. It is very possible that 6G could be the last “G.”