Introduction: When Two Innovation Speeds Collide
Artificial intelligence and mobile networks are on a collision course—but they move at very different speeds. AI technology evolves at a pace measured in months, while mobile network standards take nearly a decade from conception to commercial rollout.
This mismatch has become one of the defining challenges of 6G, the next generation of wireless technology expected to succeed 5G in the 2030s. The question facing the telecom industry is deceptively simple yet deeply complex:
How do you design a global mobile standard around AI when no one knows what AI will look like 10 years from now?
This dilemma is top of mind for industry veterans like Peter Merz, Head of Standardization at Nokia, who has spent almost 20 years shaping mobile standards. With 2025 marking a formal turning point for 6G, the industry is now confronting this challenge head-on.
2025: A Landmark Year for 6G Development
In 2025, 3GPP (3rd Generation Partnership Project)—the global body responsible for cellular standards—officially moved 6G into its study phase. This phase is expected to last 12 to 18 months and represents the first structured step toward defining what 6G will become.
During the study phase, stakeholders focus on:
Identifying future use cases
Defining performance targets
Exploring architectural concepts
Evaluating enabling technologies, including AI
Importantly, this phase does not lock in technical details. Instead, it asks fundamental questions: What problems should 6G solve? and What capabilities will future societies need from wireless networks?
Why Telecom Standards Take So Long
To understand the challenge, it helps to understand how telecom standards are built.
Mobile standards typically follow a long, consensus-driven cycle:
Research and vision building
Multi-year studies and simulations
Technical negotiations between hundreds of companies
Final specification and compliance testing
Historically:
4G LTE took roughly 10 years from early research to global deployment
5G followed a similar timeline
6G, now in studies, is expected to launch commercially around 2033–2035
This slow pace ensures interoperability, reliability, and global scale—but it clashes sharply with AI’s rapid evolution.
AI Moves at Warp Speed
Unlike telecom standards, AI innovation is driven by:
Software iteration
Massive computing power
Venture capital and cloud platforms
Open-source ecosystems
In just the past five years, AI has moved from narrow models to large foundation models, real-time multimodal systems, and edge AI deployment. Hardware accelerators, training techniques, and inference models are improving at a pace that no fixed standard can predict.
By the time 6G networks go live, today’s AI architectures may already be obsolete.
Three Roles AI Will Play in 6G
One reason the problem is so complex is that AI is not just a user of 6G—it is expected to become part of the network’s fabric.
1. AI as Traffic
AI applications generate massive volumes of data:
Distributed model training
Continuous sensor feeds
High-resolution video for perception systems
Real-time feedback loops
These workloads demand:
Ultra-high data rates
Extremely low latency
Consistent reliability
6G performance targets—such as sub-millisecond latency and terabit-per-second peak speeds—are driven largely by AI use cases.
2. AI as a Network Optimizer
AI is also expected to run inside the network itself, enabling:
Self-optimizing radio access networks
Predictive traffic management
Intelligent spectrum allocation
Automated fault detection and recovery
In this model, the network continuously learns from its environment and adapts in real time—something traditional rule-based systems struggle to achieve.
3. AI as a Native Network Capability
The most ambitious vision positions AI as a native service of the network, meaning:
Built-in support for model deployment and exchange
AI-aware scheduling and routing
Integration with edge and cloud computing
This would allow applications to request not just connectivity, but intelligence on demand.
Why Hard-Coding AI Into 6G Is Risky
According to standardization experts like Merz, the biggest risk is locking today’s AI assumptions into tomorrow’s network.
If 6G specifies:
Particular AI models
Fixed training frameworks
Narrow inference methods
…the standard could become outdated before it is even deployed.
That’s why the industry is increasingly focused on abstraction, not implementation.
Flexibility: The Core Design Principle of 6G
Rather than defining AI itself, 6G is expected to define enablers, such as:
Programmable interfaces
Modular and software-driven architecture
Support for distributed intelligence across devices, edge, and cloud
AI-agnostic frameworks that can evolve post-deployment
This approach mirrors how modern computing platforms stay relevant: through upgradability, not rigidity.
Performance Targets Driven by AI Demand
Although still under discussion, early 6G targets include:
Up to 1 Tbps peak data rates
Latency below 0.1 milliseconds
10× energy efficiency improvement over 5G
Native support for sensing, XR, digital twins, and AI workloads
These ambitious goals reflect the belief that future AI-driven services will demand far more than 5G can deliver.
What This Means for Operators and Consumers
For operators, 6G must be flexible enough to justify long-term investment while remaining adaptable to unknown AI developments.
For consumers, the stakes are tangible:
More responsive AI assistants
Seamless augmented and mixed reality
Smarter cities and transportation systems
Resilient networks during emergencies
A poorly designed standard could limit innovation for decades.
What Happens After the Study Phase
Once the study phase ends—likely in 2026—6G will enter:
Formal specification work
Technical requirements definition
Global consensus-building
From there, it will take several more years before hardware, devices, and networks begin to materialize.
By then, AI will have evolved again—reinforcing why adaptability is central to 6G’s vision.
Conclusion: Designing for an Unknowable Future
The challenge of aligning 6G with AI is not about predicting the future accurately. It is about building a system capable of evolving alongside it.
As Nokia’s Peter Merz and other industry leaders emphasize, the true innovation of 6G may not be raw speed—but its ability to remain relevant in a world where AI never stands still.
In that sense, 6G is less about technology—and more about future-proofing intelligence itself.
