Wireless networks have always advanced on a familiar rhythm. Every ten years or so, the industry has introduced a new generation – 3G, then 4G, then 5G. Each transition brought major performance improvements, but also required extensive infrastructure upgrades, long deployment cycles, and years of recovery on prices and investment.

That model worked when new generations of technology evolved gradually. But it’s become quite obvious over the last 18 months that artificial intelligence – like cloud-based software in general – does not evolve gradually. 

New AI models are arriving every few months, along with new capabilities networks will need to support. That creates a fundamental mismatch: A network architecture built around periodic hardware refreshes may struggle to keep pace with technologies that evolve continuously.

This means the most important 6G debate is not about how fast data will be able to move across networks. It is about architecture – specifically, whether 6G is built as an open, software-driven platform that can evolve alongside AI, or one tied closely to a single technology roadmap, supplier, or stack.

Those are the two distinct visions taking shape: One treats 6G as the next major upgrade cycle, delivering new capabilities through large-scale, coordinated hardware overhauls every ten years. The other treats 6G as a software-driven platform, evolving continuously through updates, automation, and open interfaces.

I believe the second path is the right one.

Think about how modern technology improves. Your smartphone gains new capabilities through software updates, not by constantly replacing it. Wireless networks should work the same way.

As AI becomes embedded in communications infrastructure, the ability to introduce new capabilities through software – rather than repeated hardware cycles – becomes a clear advantage. It allows networks to keep pace with change instead of chasing it.

Perhaps more importantly, architecture determines not just the speed of innovation, but whether it’s driven by a diverse ecosystem or a narrow set of platforms. 

Open systems ensure that new ideas, new applications, and new technologies can be introduced by many players – not just delivered according to one single roadmap. They allow operators to adopt advances as they emerge, rather than waiting for them to be bundled into the next major release cycle. And they also create a more resilient and secure ecosystem, because when no single company or platform controls the entire stack, vulnerabilities are less likely to cascade through the system.

Many of the building blocks of a more flexible, software-driven future already exist. Cloud-native networks, AI-driven automation, and open interfaces are already being deployed today. But unfortunately, there appears to be growing interest in more vertically integrated approaches, where compute, software, and network capabilities are closely coupled and optimized together.

I understand those approaches – they may offer attractive, near-term gains in efficiency or performance. But they also create bottlenecks, concentrating influence over how the network evolves and who gets to shape that evolution. Over time, that concentration can reduce the competitive pressure that drives better technologies and faster deployment. That’s why we must be cautious about any model that concentrates too much influence in any one layer of the system.

Public policy has an important role in determining which vision of 6G prevails, because the decisions being made today on spectrum and standards will help determine whether 6G develops as an open, software-driven platform or a more tightly controlled one.

Policymakers should approach those decisions with a simple principle in mind: preserve openness and flexibility. Standards should support interoperability and avoid unnecessary complexity; spectrum policy shouldn’t limit an entire generation of wireless networking to a single frequency. Industry and government must work together to ensure that no single company, platform, or technology stack becomes a gatekeeper for future innovation.

Ultimately, the most consequential 6G decisions will be made long before the first 6G network is deployed. The choices made on spectrum, standards, and the broader industry vision for 6G will help determine whether the next generation of wireless remains open, adaptable, and responsive to change.

We must get those choices right.