Traditional Telecom Operators in the AI Era | Future & AI Infrastructure

1. The AI Era Has Arrived—Telecom Operators Stand at a New Crossroads

Over the past two decades, the Internet transformed the communications industry.

Over the next two decades, artificial intelligence (AI) will redefine the entire digital world.

From ChatGPT, Gemini, and Claude to autonomous driving, humanoid robots, AI agents, and enterprise foundation models, the world is entering an unprecedented wave of AI infrastructure investment.

According to IDC, the global AI market is expected to reach several trillion dollars by 2030. However, the true foundation of AI is not the models themselves, but the infrastructure that powers them:

• GPU computing clusters

• Hyperscale AI data centers

• High-speed optical networks

• Cloud computing platforms

• Edge computing

• Energy and power systems

In essence, AI represents an infrastructure revolution.

Telecom operators occupy one of the most strategic positions within this transformation.

However, the reality is far from optimistic.

Traditional telecom operators worldwide are facing common challenges:

• Declining Average Revenue Per User (ARPU)

• Network traffic growing much faster than revenue

• OTT platforms such as WhatsApp, WeChat, Netflix, and TikTok eroding traditional telecom services

• Massive 5G investments with long return cycles

• Increasing customer resistance to paying more for connectivity

In short:

Networks are becoming more critical than ever, while monetizing them is becoming increasingly difficult.

2. The Biggest Threat Is Not AI—It's Becoming a "Bit Pipe"

The telecom industry has long discussed the concept of operators becoming nothing more than a "Bit Pipe."

What does this mean?

Internet companies create ecosystems.

AI companies create intelligence.

Cloud providers deliver computing platforms.

Telecom operators merely transport data.

Every day, users interact with:

• ChatGPT

• TikTok

• YouTube

• Netflix

• AWS

• Microsoft Azure

Most profits flow toward platform companies.

Meanwhile, telecom operators continue investing in:

• Fiber infrastructure

• Mobile base stations

• Submarine cable systems

• Network maintenance

• Power consumption

• Optical network upgrades

• Data center interconnection

Capital expenditures (CAPEX) continue to rise, while profit margins continue to shrink.

AI itself will not replace telecom operators.

The real danger is remaining a bandwidth provider instead of becoming a value-added digital infrastructure provider.

3. The More Powerful AI Becomes, the More Important Networks Become

Contrary to popular belief, AI does not reduce network demand—it dramatically increases it.

A single ChatGPT request typically involves:

User
   ↓
Access Network
   ↓
Operator Backbone
   ↓
AI Data Center
   ↓
GPU Cluster
   ↓
AI Inference
   ↓
Response Returned

Every generated token requires continuous data exchange.

AI training is even more demanding.

A hyperscale GPU cluster containing over 10,000 GPUs may generate petabytes—or even exabytes—of internal network traffic every day.

As AI adoption accelerates, future networks will require:

• Higher bandwidth

• Lower latency

• Greater reliability

• Smarter traffic scheduling

• Lossless transport

Over the next several years, AI Fabric networks will undergo their largest upgrade cycle in history.

For telecom operators, balancing traditional telecommunications services with massive investments in AI networking infrastructure will become one of the industry's greatest strategic challenges.

4. What Core Assets Do Telecom Operators Still Possess?

Although Internet companies dominate software and AI companies lead model development, telecom operators still possess strategic assets that are difficult to replicate.

Global Network Infrastructure

Operators control:

• Metropolitan Area Networks (MAN)

• National backbone networks

• International submarine cable systems

• Fiber-optic infrastructure

• 5G networks

• Future 6G networks

Regardless of how advanced AI becomes, it still depends on reliable communication networks.

Massive Data Connectivity

Telecom operators connect:

• Hundreds of millions of consumers

• Millions of enterprises

• Billions of IoT devices

Future autonomous vehicles, industrial robots, drones, and smart cities will all rely on operator-managed connectivity.

No other industry possesses such extensive network reach.

Edge Computing Infrastructure

Future AI inference will increasingly occur at the edge rather than exclusively in centralized cloud data centers.

Applications including:

• Autonomous driving

• Smart manufacturing

• Industrial automation

• AR/VR

• Robotics

require latency between 1–10 milliseconds.

Telecom operators already operate thousands of distributed facilities, making them ideal locations for:

• Multi-access Edge Computing (MEC)

• Edge AI

• Edge GPU clusters

• Regional AI data centers

AI Data Centers and GPU Cloud Resources

Operators worldwide are accelerating investments in:

• AI Data Centers

• GPU Cloud platforms

• Intelligent Computing Centers

Major operators including China Mobile, China Telecom, China Unicom, and international telecom providers are building hyperscale GPU clusters.

The future telecom operator will provide not only connectivity—but also computing power.

5. The Greatest Opportunity: Becoming an AI Infrastructure Provider

The biggest transformation is not 6G.

It is a transformation of identity.

Yesterday:

Network Provider

Tomorrow:

AI Infrastructure Provider

Instead of selling only connectivity, operators will deliver integrated infrastructure that combines:

• High-speed networking

• GPU computing

• Storage

• AI data centers

• Cloud platforms

• AI development environments

• Cybersecurity

• Data services

This evolution resembles the successful transformation of AWS from cloud hosting into comprehensive digital infrastructure.

However, the required investment will also be unprecedented.

6. Five Strategic Directions for Telecom Operators

6.1 Build AI Computing Centers

Future competition will not be determined by the number of cellular towers.

It will depend on who owns the largest AI computing infrastructure.

Key investment areas include:

• NVIDIA GPU clusters

• Liquid-cooled AI data centers

• High-speed optical interconnects

• AI training platforms

Operators may become regional AI computing providers.

6.2 Deploy AI Networks

AI workloads differ significantly from conventional Internet traffic.

Modern AI infrastructure requires:

• RoCE

• InfiniBand

• 400G Ethernet

• 800G Ethernet

• 1.6T Optical Networking

• Ultra-low latency

• Lossless networking

Operators will gradually deploy AI backbone networks, AI metro networks, and AI edge networks.

6.3 Expand Edge AI

Many AI applications cannot tolerate cloud latency.

Examples include:

• Autonomous driving

• Industrial automation

• Smart factories

• Robotics

• AR/VR

Edge computing infrastructure will regain strategic importance.

6.4 Deliver AI Industry Solutions

Future telecom services will extend beyond connectivity.

Operators can provide integrated AI solutions for:

• Healthcare

• Manufacturing

• Education

• Finance

• Smart cities

These solutions combine:

• Networks

• GPU computing

• Industry-specific AI models

• Cloud platforms

• Data services

• Cybersecurity

6.5 Embrace AI Internally

AI will reshape telecom operations themselves.

Examples include:

Ultimately, operators will evolve toward fully Autonomous Networks.

7. Optical Communications Will Be One of the Biggest Beneficiaries

AI infrastructure dramatically increases demand for optical networking.

Future telecom networks will require:

• More 400G QSFP112 LR4/FR4/DR4/SR4 Optical Transceiver丨C-LIGHT

• More 800G OSFP MSA SR8/DR8/2FR4 Optical Transceiver丨C-LIGHT

• More 1.6T OSFP-RHS DR8 & 2FR4 Optical Transceiver | AI Data Center丨C-LIGHT

• More DWDM 10G/2.5G/1G/100M丨C-LIGHT

• More DCO 800G/400G/200G/100G DWDM Coherent Transceiver丨C-LIGHT

• More Data Center Interconnection (DCI)

• More metro optical network upgrades

As a result, the optical communications industry is entering another growth cycle, including:

• Optical Transceivers  Professional Fiber Optic Products Manufacturer丨C-LIGHT

• Active Optical Cables (AOC)  800G/400G/200G/100G/50G/40G/25G/10G AOC丨C-LIGHT

• Direct Attach Copper (DAC)  800G/400G/200G/100G/50G/40G/25G/10G DAC Cable丨C-LIGHT

• Active Electrical Cables (AEC) 800G AEC QSFP-DD/OSFP cable丨C-LIGHT

• DWDM systems DWDM MUX,DWDM DeMUX丨C-LIGHT

• Optical Transport Networks (OTN) WDM/Wavelength Division Multiplexing 丨OTN/Optical Transport Network丨C-LIGHT

• ROADM

• Co-Packaged Optics (CPO)

• Silicon Photonics

For high-speed optical interconnect solution providers such as C-LIGHT, telecom operators' AI network upgrades are expected to generate sustained demand for:

• 400G Optical Transceivers

• 800G Optical Transceivers

• 1.6T Optical Transceivers

• AI Cluster Optical Networks

• Data Center Interconnection (DCI)

• High-speed optical cabling solutions

8. Who Will Be Telecom Operators' Real Competitors?

In the coming decade, telecom operators will increasingly compete with AI infrastructure providers rather than traditional carriers.

Their future competitors include:

• AWS

• Microsoft Azure

• Google Cloud

• Oracle Cloud

• NVIDIA DGX Cloud

• OpenAI ecosystem

• AI infrastructure service providers

Enterprise customers will purchase integrated solutions—not merely bandwidth.

These solutions include:

• Computing power

• Cloud services

• AI platforms

• Data services

• Secure networking

Operators must participate in this market or risk becoming commoditized connectivity providers.

9. Conclusion

AI will not eliminate telecom operators.

However, it will eliminate outdated telecom business models.

The greatest risk facing operators is not technological disruption—it is continuing to rely on revenue generated solely from bandwidth and connectivity.

Over the next decade, the world's most competitive telecom companies will evolve into comprehensive digital infrastructure providers integrating:

• Networks

• GPU computing

• AI platforms

• Data centers

• Cloud services

• Edge computing

• Intelligent data services

The communications industry is moving from connecting people to connecting intelligence.

Future network value will no longer be defined only by bandwidth.

Instead, it will be determined by a network's ability to support AI training, AI inference, real-time collaboration, and digital transformation across every industry.

For telecom operators, this represents both the greatest challenge and the greatest strategic opportunity in decades.

Those that successfully transform from Telecom Operators into AI Infrastructure Providers will be well positioned to become the foundational infrastructure companies of the AI economy.

10. Frequently Asked Questions (FAQ)