Artificial intelligence is transforming society rapidly. However, key questions remain regarding how global energy systems can be reshaped at scale.
At London Climate Action Week, a panel discussed the AI-climate nexus, exploring how AI is involved in energy systems and global growth.
Chris Skidmore, Chair of the Climate Action Coalition, moderated the panel. He noted that governments worldwide are recognising the technology's integral role in economic growth.
Legal complexities present immediate hurdles for companies deploying these technologies. Regulatory environments vary significantly between international jurisdictions.
Mark O'Conor, MD for Sectors at DLA Piper, highlights the distinct operational challenges that companies encounter.
Mark says: "It is vendor dependence and lock-in, it is IP issues, it is cyber security, it is regulatory complexity."
He believes the primary value of artificial intelligence lies in system optimisation. Legal risk management must be coupled with climate solutions.
Mark continues: "Those who succeed are those who couple the AI solution to the climate issue in a defensible way."
Modernising physical infrastructure is vital for technology deployment. Engineering and digital systems must converge to achieve operational efficiency.
Claire Gauthier, EVP, Global Head of Utilities at Capgemini, views the current climate challenges as remarkable opportunities.
Claire says: "With the convergence of the digital world and the physical world, that's an amazing opportunity through AI to get into more efficiency."
Advanced software agents can improve capital allocation projects. These tools can optimise construction schedules for new renewable energy assets.
Claire explains: "We already have some AI agents that are able to optimise by 30% the schedule."
Managing the growing grid burden
Cooling systems represent an immense drain on electrical infrastructure. The expansion of data centres is accelerating global power demand.
Hakan Yilmaz, Chief Sustainability Officer at Carrier, says: "Air conditioning brings great comfort to indoor spaces, but it consumes a lot of electricity and it creates a major peak."
The organisation is integrating batteries and cloud connectivity into future cooling hardware. Dedicated software will autonomously calculate environmental parameters.
Hakan explains: "Every battery and system will be connected to the cloud with a dedicated AI agent that optimises the energy use."
Artificial intelligence also offers significant utility for public network operators. It can manage flexibility by shifting peak electricity demand.
Lucy Yu is the CEO of the Centre for Net Zero. This research body is part of Octopus Energy Group, a renewable energy supplier. Lucy is leading an electricity network review.
Lucy says: "There are a huge number of ways in which AI can be useful for managing the grid."
Successful field trials have demonstrated how automated algorithms can shift vehicle charging patterns. This directly reduces consumer electricity bills.
Lucy continues: "We applied an AI algorithm to their charging tariff and demonstrated we could reduce peak demand across that entire fleet by 42%."
Simulating future power network operations
Renewable generation requires sophisticated forecasting due to weather variations. Advanced software models can process correlations across multiple sites.
Lucy explains: "AI techniques are particularly beneficial for that kind of forecasting if you want to forecast wind or solar."
Operators can also utilise surrogate models. These are efficient artificial intelligence networks that replicate complex physical simulations at immense speeds.
Lucy reveals: "This can enable us to run power system simulations at many orders of magnitude higher than we can currently."
Understanding technology footprints is a primary concern for hardware developers. Infrastructure energy consumption must be balanced against systemic carbon savings.
Josh Parker, Head of Sustainability at NVIDIA, says: "We start where most companies do, looking at our footprint and the footprint of our technology."
The firm has transitioned its platform to direct-to-chip liquid cooling. This method removes the need for energy-intensive chillers.
Josh explains: "The intake temperature for our cooling systems is liquid at 45 degrees Celsius, which allows us to cool more efficiently."
He notes that global data centre consumption remains a small fraction of total worldwide usage. True benefits come from practical applications.
Scaling deployment through governance frameworks
Virtual digital twins can optimise manufacturing facilities before operations begin. A digital twin is a virtual representation of a physical asset.
Josh reveals: "They used our digital twin technology and the AI embedded in that to optimise the facility for energy consumption."
Widespread adoption requires robust corporate frameworks. Executives must establish strong oversight mechanisms to handle the massive influx of international legislation.
Mark says that structure is the defining factor for organisational success: "Governance is the key that can allow you to deploy AI at scale."
Simplifying complex workflows remains a critical hurdle for heavy asset industries. Legacy systems often prevent rapid technological scaling.
Claire stresses that corporate mindsets must evolve. Organisations need to build interdisciplinary teams to manage data effectively.
Claire says: "Unlocking the value of AI will come from that workflow piece."
Grid bottlenecks remain the primary near-term constraint for innovation. The foundation of future deployment rests entirely on grid optimisation.
Hakan argues that network resilience must be achieved before infrastructure can expand safely. Energy flexibility remains paramount.
Hakan explains: "Once we build that energy flexibility, that is the foundation to scale AI."
