As artificial intelligence increasingly relies on vast amounts of energy, the technology and energy sectors face technical, operational, and legal challenges. This article explores the mounting electricity demands, opportunities for AI-enabled energy optimisation, and the complex contractual considerations shaping the future of AI-energy collaborations.
The expanding use of artificial intelligence (AI), particularly generative AI, is creating increasingly strong connections between the technology and energy sectors, due to the substantial power consumption associated with AI systems. As the adoption of AI technologies intensifies, key challenges emerge relating not only to technical and operational constraints but also to the legal and contractual frameworks governing these developments.
A report by Computer Weekly highlights that training a single large language model can consume as much electricity as a small town, underscoring the significant energy demands involved. Currently, data centres account for approximately 1.5% of global electricity use. The International Energy Agency (IEA) forecasts that this demand will more than double by 2030, driven predominantly by AI workloads. The projected growth would require new global energy capacity equivalent to around four times the United Kingdom’s present total electricity consumption.
The increased energy requirements are primarily concentrated in regions hosting data centres. This situation places considerable strain on local power grids, necessitating substantial and rapid upgrades to grid infrastructure or compelling data centre owners and operators to swiftly secure reliable and sustainable energy sources dedicated to their facilities. The competition to access sufficient energy resources is intensifying as AI becomes more prevalent.
Paradoxically, while AI applications require large amounts of power, AI technologies also hold promising potential to improve energy management. AI can optimise power grid operations, facilitate the integration of renewable energy sources, predict equipment failures, and enhance energy efficiency across industries and buildings. Despite this potential, the energy sector has been slower to adopt AI technologies compared to industries such as technology and finance. Greater AI integration within energy management systems is anticipated in coming years.
The interplay between AI and energy brings additional complexities from a legal and contractual perspective. The frameworks governing AI-energy projects involve navigating diverse regulatory regimes, complex supply chains, and geopolitical uncertainties. This leads to intricate negotiations focusing on risk allocation, pricing structures, and responsibilities to avoid operational downtime. Moreover, the regulatory environments for both AI and energy are rapidly evolving, which complicates efforts to achieve compliance and contractual certainty.
According to Computer Weekly, it is essential during the contract drafting phase to anticipate potential disputes and implement mechanisms for their early resolution or avoidance. Contracts must be carefully constructed to foresee possible issues while maintaining sufficient flexibility to accommodate unpredictable factors in projects expected to span decades. This entails clearly defining stakeholder responsibilities, establishing measurable performance criteria with agreed monitoring methods, and effectively assigning risks between parties.
In summary, the growing interdependence of AI technologies and the energy sector presents both opportunities and challenges. The technical demands of AI necessitate robust energy infrastructure, while AI itself offers tools for more efficient energy use. On the legal front, carefully negotiated agreements and regulatory compliance will be critical to the long-term success of projects at the intersection of these dynamic industries.
Source: Noah Wire Services
- https://www.polytechnique-insights.com/en/columns/energy/generative-ai-energy-consumption-soars/ – This source corroborates the claim that training large AI models consumes significant electricity, comparable to small towns, and that AI energy use is rapidly increasing with generative AI’s rise, matching the article’s emphasis on energy demands of AI systems.
- https://www.iea.org/news/ai-is-set-to-drive-surging-electricity-demand-from-data-centres-while-offering-the-potential-to-transform-how-the-energy-sector-works – This IEA report supports the article’s information on data centres accounting for about 1.5% of global electricity consumption, the forecast that electricity demand will more than double by 2030 mainly due to AI workloads, and the need for new energy capacity on a scale comparable to multiple UK electricity consumption.
- https://e360.yale.edu/features/artificial-intelligence-climate-energy-emissions – This article confirms the projection that data centers’ electricity consumption in 2026 will double compared to 2022, matching the scale of a country such as Japan, and discusses the operational factors that create energy strain and opportunities for AI to improve energy efficiency.
- https://mitsloan.mit.edu/ideas-made-to-matter/ai-has-high-data-center-energy-costs-there-are-solutions – This MIT Sloan article verifies the challenges AI data centers pose to energy infrastructure, the share of global energy demand data centers consume, and also discusses AI’s potential for energy optimization and efficiency improvements, supporting both the energy impact and the potential AI benefits highlighted in the article.
- https://www.statista.com/statistics/1536969/ai-electricity-consumption-worldwide/ – This source provides statistical data on AI’s electricity consumption globally, supporting claims about current power usage, its proportion relative to total data center consumption, and forecasts for significant growth by 2028, consistent with the article’s projections on AI energy demand.
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
8
Notes:
The narrative discusses current trends concerning AI energy consumption and references recent forecasts by the International Energy Agency projecting demands up to 2030, indicating up-to-date information. No references to outdated roles or events are present and the data aligns with ongoing discussions about AI and energy usage. There is no explicit indication that this is a press release; thus, freshness rating is high but not perfect due to lack of direct publication date or source timestamp. The mention of ‘Computer Weekly’ report implies the data originates from a known contemporary source.
Quotes check
Score:
7
Notes:
The narrative contains paraphrased information from Computer Weekly but no direct verbatim quotes requiring verification. The mention that training a large language model consumes as much electricity as a small town is a widely cited claim in the AI energy discourse, though the exact original source and earliest date of this claim are not pinpointed here. Since no exact quotations are used, and the claim is consistent with public knowledge, the score is moderately high.
Source reliability
Score:
6
Notes:
The narrative references Computer Weekly and the International Energy Agency (IEA), both recognised organisations. Computer Weekly is a known technology publication but less prestigious than major global outlets like BBC or Reuters. IEA is highly credible for energy data. The lack of direct links or clear publication information reduces confidence slightly. Overall, the narrative is likely reliable but not from the most authoritative journalistic brand.
Plausability check
Score:
9
Notes:
Claims about AI’s energy consumption and its doubling by 2030 align with well-documented industry estimates. The dual role of AI as a power consumer and a tool for energy optimisation is consistent with current expert understanding. The challenges in legal and contractual frameworks around AI-energy projects are plausible given regulatory complexities in both sectors. No contradictory or implausible claims appear. Verification of precise figures is limited but the overall assertions are credible and fit ongoing technological trends.
Overall assessment
Verdict (FAIL, OPEN, PASS): PASS
Confidence (LOW, MEDIUM, HIGH): HIGH
Summary:
The narrative provides a fresh and plausible overview of the relationship between AI and the energy sector, referencing widely accepted forecasts and known challenges. It draws on recognized organisations and presents generally trustworthy information without outdated data or unverifiable quotations. While the narrative’s source is not from a top-tier global news outlet, the information aligns with credible contemporary discussions, supporting a high-confidence pass verdict.
