ENERGY SUPPLIERS EMBRACE ARTIFICIAL INTELLIGENCE (AI)
AI is very much top of mind across all constituencies, largely amplifying the effectiveness of existing operations through data analysis, demand/load management, digital twinning, and efficiency applications.
Last year, energy suppliers surveyed saw numerous barriers when it came to AI adoption including capital expense, workforce challenges, cybersecurity concerns, and technology development. While those same concerns exist today, they seem to have been eclipsed by the potential of AI. When asked “specifically with regard to your business, which of the following do you feel have the most long-term promise”, AI ranked as top answer by 44% of supplier and investor respondents with the U.S. and the Middle East showing the greatest level of interest at 56% each.
Currently, 45% of respondents are already using and investing in AI technologies, while an additional 40% are considering adoption. This trend is particularly strong in the Middle East (55%), the U.S., and Latin America (both at 48%). The focus has shifted from overcoming adoption hurdles to leveraging AI for operational optimization. As AI proliferates, the talent gap in that area is sure to widen. 33% of survey respondents, especially in the U.S. (47%) and Middle East (45%), suggest that digital and AI driven skills will be most in demand over the next five years.
It is worth noting here that the U.K. and the EU appear to lag in AI investment, perhaps due to the EU’s adoption of the AI Act on August 1, 2024. An early effort to legislate and regulate the use of artificial intelligence, the act seeks to mitigate risk and provide regulatory guidance for “providers, deployers, importers, distributors, representatives, and affected persons." Similarly, the new U.K. Labour government has announced plans to establish legislation for “the most powerful artificial intelligence models.”
Which best defines your organization's status when it comes to using or investing in AI for energy-related issues? (energy suppliers and investors)
Executives primarily see AI's value in enhancing existing operations. Data mining and knowledge discovery lead at 48%, closely followed by maintenance (47%), demand/load management (42%), and operating controls (41%). However, regional priorities show significant variation. The EU is taking a broad approach to AI implementation across operations, while the U.S. is particularly interested in resource discovery and optimizing new technologies like carbon capture. The Middle East is focusing on smart buildings and digital twinning, the U.K. on data mining and operational controls, and Latin America on maintenance (51%) and demand load management.
Commercial consumers include AI in the subset of means by which they plan to assume affordable, reliable energy supply (29%). We found those most likely to do so are in Asia-Pacific (41%, and the top choice in that region), the Middle East (44%), and the U.S. (38%).
Where is AI already having an impact on energy-related business functions, or where do you expect it to have an impact? (energy suppliers and investors)
As noted above, AI adoption still faces challenges. Cybersecurity is the primary concern globally, cited by 51% of respondents. Other key issues include data storage, privacy, power/energy requirement and data ownership. The ability to measure return on investment (ROI) is more evident in developed markets such as the U.S., U.K., and Middle East.
Which of the following present current or future challenges to energy digitization?
AI's hunger for power
The Asia-Pacific region diverges from the global trend, pointing to the power and energy requirements of AI (49%) as a slightly bigger concern than cybersecurity. The region's prioritization of AI's energy requirements over cybersecurity concerns highlights a growing awareness of AI's energy demands.
This shift in focus is supported by alarming statistics: a single AI query consumes ten times more electricity than a standard Google search, and the International Energy Agency projects that data center energy consumption will double by 2026 compared to 2022 levels. Furthermore, the CEO of ARM, a major semiconductor company, recently suggested that data centers could account for 25% of U.S. electricity consumption by 2030, a dramatic increase from the current 4%.
These projections underscore the potential for AI to become a major driver of energy demand, even as it enhances efficiency, presenting significant challenges for energy management, sustainability efforts, and infrastructure planning in the coming years. As AI adoption accelerates across various sectors, including energy, managing AI’s energy use is a quickly rising priority, necessitating innovative solutions in energy efficiency, renewable energy integration, and grid management. We have already seen one such notable response.
In March, a global online retailer purchased a nuclear-powered data center in the U.S., signaling a recognition of the need to source cleaner energy to support high energy demand technology. Microsoft has announced its intent to purchase the generating capacity of Unit 1 of Constellation Energy’s Three Mile Island nuclear plant to power its data centers for 20 years. The unit is currently expected to come online in 2028. The U.S. is already seeing the highest level of small nuclear reactor (SNR) interest across all geographies surveyed. It is likely that interest will grow as AI data center needs increase.
Regional: Which of the following present current or future challenges to energy digitization?
Bias in AI: tackling the algorithm
One significant concern not highlighted in our survey is the risk of bias in AI-generated data, which can undermine trust and lead to unintended consequences in decision-making processes. AI systems can perpetuate or amplify existing biases when trained on skewed datasets, potentially resulting in discriminatory outcomes across various sectors. For energy producers and consumers, this could influence pricing, investment decisions, and resource allocation. Further, this issue is particularly concerning in the context of energy equity, as algorithms may under represent or overlook disadvantaged communities, potentially exacerbating existing disparities in access to affordable and reliable energy resources.
The need for oversight and standards is highlighted in a recent publication from the U.S. National Institute of Standards and Technology, "Towards a Standard for Identifying and Managing Bias in Artificial Intelligence," which calls for higher ethical standards in AI development. As the use of AI continues to proliferate among energy suppliers and consumers, certainly this topic will prove to be of growing concern.
The rapid transition in mindset of energy suppliers and investors regarding the use of AI reflects a broader trend where AI is moving from an experimental technology with adoption challenges, to a key component of business operations. This suggests that the energy sector is rapidly moving into a new phase of digital transformation. The question is no longer whether to adopt AI, but rather how to best leverage it for efficiency and innovation. This evolution presents both opportunities and challenges for energy professionals who must now navigate the complexities of AI implementation while addressing persistent concerns around cyber security, data management, and bias.
Share on social:
