“With great power comes great responsibility.”

It may be a line borrowed from Spider-Man, but it captures the inflection point facing the telecoms industry today.

Telecom networks are the invisible infrastructure of modern life. Every message sent, video streamed and connected device depends on the vast digital highways that telecom operators build and maintain.

But the industry is now at a crossroads.

As UK telecoms intensify their journey toward Net Zero and with 6G on the horizon, AI is emerging as both a transformative enabler and a sustainability challenge.

A dual approach to AI

Telecoms has entered a new phase of AI-fueled growth. Data demand is surging, networks are becoming denser, and the next wave of innovation is pushing digital infrastructure to new limits.

With this growth comes a clear opportunity to lead on sustainability – but to seize it, the industry must unlock AI’s full potential responsibly. On one hand, AI enables smarter operations across the telecoms system, from network traffic optimization to supply chain analytics and carbon reporting. On the other, it brings rising energy and water demands driven by compute-intensive workloads.

A dual approach is therefore central to ensuring technological innovation supports the climate agenda. That means advancing both AI for sustainability – using AI to reduce emissions, optimize resources and accelerate progress towards Net Zero – and sustainable AI, ensuring that AI itself is designed, deployed and governed in a way that minimizes its environmental impact.

This is the defining tension of the next decade: scaling intelligence without scaling impact.

Pinpointing high-impact AI use cases

AI is already creating new possibilities for progress, but real value lies in prioritizing use cases that drive both business value and environmental benefit. Not every use case delivers equal value—and in a resource-constrained world, prioritization is critical.

First and foremost, energy optimization is an area of significant opportunity. Energy consumption accounts for up to 40% of telecoms’ network OPEX, and GSMA Intelligence indicates that operators could reduce OPEX by 4% for a 20% reduction in energy costs through power efficiencies.

Here, AI can assist with network load balancing to help distribute traffic evenly and reduce unnecessary energy consumption. Meanwhile, energy use in data centers can be reduced through intelligent workload distribution, predictive cooling and server optimization, all of which lower overall demand. In effect, AI enables networks to think more intelligently about how and when energy is used.

On the reporting side, AI is also playing an increasingly important role. It’s being used to support customer carbon reporting, particularly in analysing large datasets from multiple sources – estimating notoriously obscure Scope 3 emissions and helping customers better understand their impact.

For example, a network operator seeking to assess B2B customer emissions can integrate data from traffic records and device energy use, linking it with customer profiles to generate more accurate insights. What was once opaque is becoming measurable, and therefore manageable.

Establishing the foundations for sustainable AI

Without the right foundations, scaling AI risks putting the cart before the horse. Organizations must establish strong governance to ensure AI is used ethically and sustainably – and that requires close collaboration across sustainability, technology and operational teams.

A centralized governance body can accelerate decision-making, enforce accountability, and ensure AI initiatives are aligned with wider sustainability goals. It also provides the guardrails needed to scale AI with confidence.

This oversight should extend across the full AI lifecycle — from design and development to training and deployment — embedding environmental, social and ethical considerations from the outset, rather than treating them as an afterthought.

For instance, organizations can adopt a “decision tree” approach to evaluate whether AI is needed at all, and if so, what type is most appropriate, considering alternatives that may deliver similar outcomes with a lower environmental footprint.

Assessing and mitigating environmental impact

Understanding AI’s environmental footprint must be the starting point for responsible adoption. After all, you can’t manage what you can’t measure. In truth, only a minority of companies are actively monitoring and disclosing the environmental footprint of their AI models, and even fewer have set reduction targets. This indicates a broader need for lifecycle assessments and accountability in AI operations.

But this is a daunting task on a few levels. Many organizations lack transparency from AI providers regarding energy efficiency and carbon footprints, which is crucial for informed decision-making – and without clear, standardized metrics and greater visibility across the value chain, organizations are left making critical decisions in the dark. Addressing this gap will require closer collaboration across the ecosystem to establish common standards, improve transparency and embed sustainability into AI decision-making.

A responsibility that extends beyond telecoms

The impact of telecoms sustainability decisions goes far beyond the industry. Telecoms already account for at least 1.6% of global carbon emissions. Decisions made here influence energy demand, emissions reduction and resource efficiency across national economies. By acting decisively, telecoms can set a standard for other sectors, demonstrating that AI-powered growth and sustainability are not mutually exclusive.

But there is work to be done. As AI adoption accelerates across networks, customer operations and enterprise services, understanding and managing its energy impact will become increasingly important. Ultimately, the challenge facing telecoms is not simply to connect the world, but to do so responsibly.

We feature the best business phone systems.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit