The government is urged to mandate stricter reporting for data centres to mitigate environmental risks associated with the AI sprint.
A report published today by the National Engineering Policy Centre (NEPC) highlights the urgent need for data centres to adopt greener practices, particularly as the government’s AI Opportunities Action Plan gains traction.
Unlocking the potential of AI while minimising environmental risks
AI is hailed as capable of driving economic growth, creating jobs, and improving livelihoods. Launched as a central pillar of the UK’s tech strategy, the AI Opportunities Action Plan is intended to "boost economic growth, provide jobs for the future and improve people’s everyday lives." However, this growing reliance on AI also poses environmental risks from the infrastructure required to power these systems.
Five steps to sustainable AI
The NEPC is urging the government to spearhead change by prioritising sustainable AI development. The report outlines five key steps policymakers can act upon immediately to position the UK as a leader in resource-efficient AI:
- Expand environmental reporting mandates: Mandate data centres to report on their energy sources, water consumption, carbon emissions, and e-waste recycling practices.
- Communicate the sector’s environmental impacts: Raise public awareness of the environmental costs of AI to encourage developers to optimise AI tools, use smaller datasets, and adopt more efficient approaches.
- Set sustainability requirements for data centres: Ensure data centres are designed with environmental sustainability in mind, using renewable energy, reducing water use, and minimising carbon emissions.
- Reconsider data collection, storage, and management practices: Re-evaluate data collection, storage, and management practices to reduce unnecessary environmental costs.
- Lead by example with government investment: Demonstrate a commitment to sustainable AI development through government investment in green data centres and sustainable infrastructure.
Smarter, greener data centres
One of the most urgent calls to action involves redesigning data centres to reduce their environmental footprint. The report advocates for innovations like waste heat recovery systems, zero drinking water use for cooling, and the exclusive use of 100% carbon-free energy certificates.
Time for transparency around AI environmental risks
Public awareness of AI’s environmental toll remains low. Recent research by the Institution of Engineering and Technology (IET) found that fewer than one in six UK residents are aware of the significant environmental costs associated with AI systems.
Conclusion
As the UK pushes forward with ambitious plans to lead in AI development, ensuring environmental sustainability must take centre stage. By adopting policies and practices outlined in the NEPC report, the government can support AI growth while safeguarding finite resources for future generations.
Frequently Asked Questions
Q: What are the environmental risks associated with AI?
A: AI’s growing reliance on data centres poses significant environmental risks, including energy consumption, water usage, and carbon emissions.
Q: What are the five steps to sustainable AI development?
A: The NEPC report recommends expanding environmental reporting mandates, communicating the sector’s environmental impacts, setting sustainability requirements for data centres, reconsidering data collection, storage, and management practices, and leading by example with government investment.
Q: How can data centres reduce their environmental footprint?
A: Data centres can reduce their environmental footprint by using renewable energy, reducing water use, and minimising carbon emissions through innovations like waste heat recovery systems and zero drinking water use for cooling.
