Powering Possible 2026: Energy Resilience in the Intelligence Era

This document explores strategies and innovations for achieving energy resilience in an information-driven age by 2026.

POWERING POSSIBLE 2026 Energy Resilience in the Intelligence Era

Table of contents 01 Executive Summary 8 | 02 | From AI Ambition to AI Value | 14 | | The focus is shifting from AI adoption to AI returns | 16 | | Intelligence moves into operations | 18 | | Cybersecurity and data readiness remain challenges for scale | 20 | | Capital and policy are being drawn towards the grid | 22 | | Intelligence is getting physical | 24 | | AI’s progress depends on people | 26 | | Resilience is emerging as a major driver of AI investment | 28 | | Cooperation is key for AI governance | 30 | | Grid modernization remains key concern | 32 | 03 Key Insights and Recommendations 34 04 Methodology 36 Powering Possible 2026 5

This year's Powering Possible report shows an energy industry that has moved decisively from experimenting with AI to deploying it at scale. Investment is turning into real business value, and resilience, reliability and adaptability now matter as much as efficiency. Intelligence is fast becoming part of how energy is produced, delivered and used. That shift is opening up new possibilities. Agentic AI is helping organizations coordinate and adapt in real time. Physical intelligence, which brings AI and robotics together, is turning insight into action by maintaining critical infrastructure and keeping operations running around the clock. At ADNOC, we are seeing this first-hand as we embed AI across our operations to optimize every barrel, every megawatt and every production line. As intelligence runs deeper through the energy system, our responsibility grows with it. AI must be deployed safely, securely and with rigorous governance. And people must remain at the center, building, governing and applying AI to solve real operational challenges every day. The next chapter will be defined by those who turn AI into outcomes: stronger performance, greater resilience and the ability to adapt in an increasingly complex world. ADNOC is committed to playing its part. We will share what we learn, work with partners like Microsoft and help show what is possible when energy and intelligence advance together. His Excellency Dr. Sultan Ahmed Al Jaber Managing Director and Group CEO, ADNOC and Executive Chairman, XRG

Powering Possible 2026 captures an important moment in the evolution of artificial intelligence. As AI becomes more powerful and more widely used, its relationship with energy is becoming increasingly important. Each will play a critical role in enabling the other, and together they will help shape the next era of economic growth and technological progress. This year’s findings also show how quickly the conversation is maturing. The question is increasingly moving beyond whether organizations will adopt AI to how they can use it to create lasting value. Across the energy sector, AI is already helping organizations operate more efficiently, make better-informed decisions, and get more from existing infrastructure. The opportunity now is to build on those gains and think more broadly about how AI can strengthen resilience in a world that is becoming more complex. Trust will be fundamental to that progress. As AI becomes more deeply integrated into organizations and increasingly capable of taking action, people need confidence in how these systems are governed. Innovation and responsibility need to advance together, with people remaining at the center of how AI is developed and used. What gives me optimism is the growing recognition that realizing AI’s potential will depend on more than technology alone. It will require continued investment in the infrastructure and skills needed to support it, along with collaboration across industries and borders. The opportunity ahead is significant. By bringing together advances in energy and artificial intelligence, while building the trust and human expertise needed to support them, we can help create stronger, more resilient economies and new opportunities for people around the world. Brad Smith Vice Chair & President, Microsoft

Leaders are confident of AI’s value AI investment continues to accelerate as organizations move from experimentation toward measurable returns. % Increased AI 89 investment over the past year Grid modernization remains key concern The grid is still the leading infrastructure priority as AI increases pressure on energy systems. % Prioritize grid 68 modernization to support AI growth The same challenges remain Cybersecurity, data quality and talent are the leading barriers to AI adoption. % Cite cybersecurity 43 as a barrier to AI adoption AI growth is reshaping energy demand Leaders expect AI growth to increase energy consumption, strengthening the case for energy infrastructure investment. % Expect AI to increase 74 national energy consumption by 2030 Agentic AI is scaling Integrated agentic AI is moving rapidly into organizations as adoption matures beyond pilots. Agentic AI % deployment has 33 more than doubled, up from 15% from 2025 Broader AI benefits gain recognition Supply stability and reliability now outrank prediction accuracy and cost savings as measures of AI success. Rank supply stability % and reliability as the 39 leading measure of AI’s contribution Intelligence is getting physical Autonomous and robotic systems are moving into real operational environments. Industry respondents are % deploying autonomous 60 and robotic systems in physical operations Energy supply is harder to predict Growing complexity is strengthening the focus on adaptability, reliability and resilience. Say global energy % supply has become 86 more complex and harder to predict POWERING POSSIBLE 2026 AT A GLANCE 8

Last year’s Powering Possible captured the energy sector moving AI from promise to practice. In 2026, that momentum has accelerated. Eighty-nine percent of energy leaders have increased their AI investment over the past year, while 26% now rank AI as their organization’s highest priority, up from 9% in 2025. Nearly two-thirds place it among their top two priorities. That commitment is translating into action. The use of integrated agentic AI has more than doubled, from 15% in 2025 to 33% in 2026, as organizations move beyond isolated applications and embed AI more deeply into their operations. At the same time, expectations of AI are growing. Value is moving beyond efficiency and cost savings towards stronger, more reliable and resilient operations. And as AI drives greater demand for energy, it can also help energy systems become more flexible, more reliable and more resilient. But the next phase of growth is likely to be harder, and the value of AI will not be measured only by productivity, cost reduction or the number of use cases deployed. The global energy system is Executive Summary Powering Possible began in 2024 to explore what AI would mean for energy and, in turn, what energy would mean for AI. Two years on, those questions are becoming operational as AI scales across the sector and agentic workflows and physical systems move into the field. Organizations are beginning to see returns on investment, while also understanding how AI can strengthen adaptability and resilience in a more disruption-prone world. AI and the New Value Equation becoming more exposed to disruption, from extreme weather and geopolitical volatility to growing complexity across supply chains, infrastructure and electricity networks. A resilient and adaptable global energy supply has never been more important, and AI has a leading role to play. Resilience enters the equation Eighty-six percent of energy leaders say global energy supply has become more complex and harder to predict over the past five years, and 60% expect that to be permanent. Across the energy leaders and global markets surveyed, the direction is broadly consistent, but there are noticeable differences in pace, priorities and pressures between markets. What’s clear from the data is that organizations are investing not only in the intelligence needed to anticipate change, optimize operations and respond faster when disruption occurs. AI is increasingly seen as a critical enabler of a more adaptive, integrated and resilient energy system. When asked to name the greatest strategic risk to energy security this decade, over half point to how well the system is integrated: its ability to connect infrastructure, markets and geographies and absorb shocks. For years, the industry has measured value in terms of cost, productivity and output. AI can improve all three, but this year’s findings suggest leaders are widening the lens. Nearly two-thirds say greater use of AI makes energy supply more resilient, and 82% expect AI to support and improve energy resilience by 2050. That matters because of the operating conditions the sector now faces. AI for energy, energy for AI That operational shift is changing how leaders define value. Energy supply stability and reliability is now the leading measure of AI’s contribution, named by 39% of leaders, up from 26% in 2025. Over the same period, the share prioritizing prediction accuracy and data use has fallen from 24% to 19%, while cost savings has declined from 19% to 13%. Leaders are increasingly judging AI not only by the performance of an individual model or use case, but by the performance of the energy system around it. Powering Possible 2026 9

Figure 1 Almost two-thirds of energy leaders view AI as a top organisational priority AI as a corporate priority, 2024 - 2026 | 2024 | 14% | 31% | 47% | 7% | 2% | | 45% | | 2025 | 9% | 33% | 49% | 8% | 2% | | 42% | | 2026 | 26% | 36% | 34% | 3% | 62% The top priority One of the top two priorities Don’t Know / Not Relevant One of many priorities Not a priority While supply volatility and disruption Realizing those benefits will “Energy and artificial is increasing, the system is also depend on the foundations around intelligence are being asked to carry more, and the the technology: sufficient power, becoming increasingly industry remains split on whether trusted governance, common interconnected, with infrastructure can cope. The IEA standards, skilled people and each playing a vital role in projects that electricity demand from infrastructure capable of supporting shaping the future of the data centers will reach roughly 950 a more intelligent energy system. other. Realizing their full TWh by 20301, arriving on networks potential requires an Investments into returns that must also absorb electrification integrated approach that and rising demand. Just over half of The economics are also becoming aligns energy and digital leaders are confident that existing more operational. Falling costs are infrastructure, advances infrastructure can meet AI-driven making AI more accessible, but innovation, and invests electricity demand through 2030. agentic systems and continuous in human capabilities. Forty-seven percent are not. Little workloads are increasing By strengthening this has changed on that question over consumption. As AI moves into connection, we can the past year. production, organizations are under build more secure, growing pressure to justify recurring efficient, and resilient Reliable power for AI depends on spend and secure future investment. energy systems while more than generation and electricity Adoption is no longer enough. Value supporting sustainable networks. Gas-fired capacity can must be measured through better economic growth.” support growing, concentrated decisions, improved workflows, demand only when production, stronger resilience and reduced H.E. Suhail Mohamed processing, transportation and operational risk. Al Mazrouei storage are developed in step with The early evidence on returns is Minister of Energy power generation and the grid. The and Infrastructure, encouraging. One in five leaders say United Arab Emirates resilience challenge is therefore one AI investments already made have of integration across the full energy- exceeded their costs, while 63% delivery chain. expect them at least to break even. This heightens the importance of Confidence is even stronger looking how the sector should judge AI’s ahead: nearly nine in ten expect value. Beyond productivity, AI can future investments to cover or help optimize assets, forecast risk, exceed costs, and almost half expect coordinate power flows and returns within three years. strengthen the resilience of supply. 1 Key Questions on Energy and AI, World Energy Outlook Special Report, IEA, April 2026, https://www.iea.org/reports/key-questions-on-energy-and-ai 10

The research suggests leaders divided over whether that system “The benefits of understand both the opportunity is ready to support AI at scale. intelligence must spread and what is required to capture that The challenge for energy leaders as widely as its influence. value. They see the greatest long- is no longer simply generating Responsible adoption term value of AI not at the level of a and diffusion of AI takes and delivering energy efficiently. single asset or enterprise, but in trusted infrastructure, It is building systems that can flex, orchestration across organizations reroute, optimize and adapt in skilled people and clear and borders. They identify grid real time. The organizations that human accountability. In modernization as the leading succeed will combine physical energy, those foundations investment priority to support turn AI advances into infrastructure with digital AI growth. They place security, intelligence, using AI to move more reliable supply and reliability, transparency and common faster than disruption rather than greater resilience. For the technical standards at the center UAE and G42, leadership merely reacting to it. That will of AI governance. They are also require trusted data exchange, means working with investing in the skills needed to work clear authority, interoperable partners like ADNOC to with more autonomous systems, systems, physical networks and help countries build these while recognizing that accountability hardware that can turn intelligence capabilities around their cannot be automated. own priorities. A country's into operational value. Building the system starting point should not Expectations of AI are therefore The findings in this year’s report determine its place in the describe a sector that has settled intelligence era.” rising faster than confidence in the the question of whether AI belongs | infrastructure on which it depends. | in energy and moved on to the | Peng Xiao | | Fifty-nine percent now expect AI to | harder work of making it secure | CEO, G42 | | have a large impact on grid capacity | | and dependable. The next phase | | and reliability, up from 45% in 2025. | of growth will belong to the | | Yet confidence that existing infrastructure can meet AI-driven | organizations and countries that | | can bring these requirements | | electricity demand remains | | together. Better models will matter, | | unchanged at 52%. This is one of the | | but stronger systems will matter | | report’s central tensions: leaders | | more: the power, trust, standards, | | increasingly expect AI to strengthen | | people and resilience needed to | | the energy system, while remaining | | make intelligence valuable at scale. | | Figure 2 | | Energy leaders still expect AI to increase | | the speed at which the energy sector | | becomes more sustainable | | Impact of AI adoption on sustainability | | 2% | 4% | 2% | 2% | 2% | | 10% | 6%6% | 6% | 5% | 12% | 10% | 18% | 14% | | 4% | 14% | | 17% | | 20% | 22% | 20% | 27% | | 71% | 88% | 85% | 79% | 66% | 65% | 62% | 59% | 2026 overall India Brazil China Japan US UK Germany Increase the speed Have no impact Decrease the speed Don’t know Powering Possible 2026 11

• Standards make coordination possible. The connections between assets, companies and markets warrant deliberate investment. Shared standards, trusted data exchange and agreements for coordinated action should form part of building a more capable energy system. • Accountability remains with people. Progress should be assessed through the work people and intelligent systems can complete responsibly together. Agentic workflows and physical intelligence will require usable data, skilled teams, clear authority and human oversight where decisions carry operational consequence. • Resilience means keeping options open. The ambition should be to preserve reliable supply as conditions change: seeing pressure earlier, retaining workable alternatives and responding with greater confidence. AI business cases should make these capabilities visible alongside financial and operational returns. The report points to a clear conclusion: AI creates the most value when it expands the energy sector’s capacity to adapt. That value depends on better models and more use cases, but also on the practical requirements that allow AI to operate safely, reliably and at scale. The findings point to five priorities: • Power and intelligence should be built together. Fuel supply, transportation and storage, generation, electricity networks and digital capability should be planned and financed as a connected investment agenda. This would link the dependable energy required by AI with the intelligence intended to strengthen energy supply. • Trust has to be designed into the system. Security, reliability, transparency and accountability should be treated as operating requirements, not afterthoughts. As AI begins to coordinate decisions across assets and borders, confidence will depend on knowing how systems act, who has authority and where responsibility sits. What AI requires to scale “As AI increases demand for power and economies continue to electrify, energy systems will be asked to do more than ever before. Meeting that challenge is not simply a question of adding capacity. It is about building a system that can adapt, integrate new technologies and respond to changing conditions with confidence. Intelligence can help us unlock more value from existing infrastructure while guiding where future investment is needed. Ultimately, resilience will come not from any single technology, but from how effectively we bring energy, infrastructure and innovation together to power the next era of growth.” Mohamed Al Ramahi CEO, Masdar 12

“AI gives us an opportunity to unlock much more value from the expertise, data and technology we already have. At bp, we are seeing how putting better insights into the hands of experienced engineers and operators can improve decisions and performance. Ultimately, AI will be judged by what it delivers: safer, more reliable and more efficient energy.” Meg O'Neill CEO, bp Figure 3 The majority of energy leaders view AI as a force for good Impact of artificial intelligence | 4% | 6% | 6% | 7% | 6% | 6% | 6% | | 13% | 2% | 19% | | 17% | 16% | 22% | 83% 94% 94% 90% 81% 77% 78% 72% 2026 overall India UK China Brazil US Germany Japan AI is a force for good AI causes more harm Don’t know Powering Possible 2026 13

AMB AI V

FROM AI BITION TO VALUE KEY FINDINGS 2026

Energy leaders are no longer asking whether AI has a role to play. They are treating it as a strategic priority and focusing on the value it can deliver. Nearly nine in ten organizations increased their spending on AI and digital infrastructure over the past twelve months, 44% of them significantly. AI now sits at the top of the corporate agenda: it is the single highest priority for 26% of leaders and one of the top two for a further 36%; barely 3% say it is not a priority at all. That reverses a decline. In 2024, 14% named AI their top priority; by 2025 that had fallen to 9%, as organizations turned from the excitement of the technology to the practicalities of integrating it. The investment case for AI is settled. Energy leaders are now focused on where value will emerge and how to turn AI investments into measurable performance gains, greater resilience and competitive advantage. The conversation has shifted again. It is no longer centered on whether AI works but on where it creates the most value and how that value is measured. Early returns are encouraging. Almost two-thirds of leaders believe existing AI investments will generate returns that will either cover or exceed their costs with one in five leaders already expecting strong returns. Nearly half expect their own organization to see returns within three years. These findings are in line with wider industry reports looking at the return on investments from AI. Deloitte’s 2025 survey of executives across Europe and the Middle East, looking at AI spend and returns paints a similar picture – 85% increased investments over a twelve-month period, and the majority reported achieving satisfactory ROI on a typical AI use case within two to four years2. However, it is also undeniable that fewer leaders can point to realized returns today than expect them tomorrow. It is also true that returns on AI spend are taking longer to materialize, on average, when compared to other technology investments (typically 7-12 months according to a study commissioned by AWS which looked at targeted payback timescales for major IT investments)3. Leaders report AI investments are covering costs with future spend expected to give stronger returns Figure 4 54% 35% 7% 3% 2% InvestmentExisting 21% 42% 28% 7% 3% Future Investment Clearly positive ROI Likely to break even Likely negative ROI Clearly negative ROI Don’t Know Perceived return on investment (ROI) from AI 1 THE FOCUS IS SHIFTING FROM AI ADOPTION TO AI RETURNS 2 AI ROI: The paradox of rising investment and elusive returns, Deloitte, October 2025 3 Evaluating the ROI of Major Tech Investments, Enterprise Strategy Group (commissioned by AWS), October 2024 16

AI economics create a second reason for scrutiny. A lower cost per token does not automatically mean lower total spend; it can widen usage. Every additional agent step, retrieval call or extended context carries a marginal cost, alongside the power, data and governance needed to support it. The commercial test is therefore not just whether AI works, but which use cases deserve more tokens, more autonomy and more infrastructure because they produce measurable operational or resilience value. The data suggests that the industry is investing not only on evidence, but also on conviction: a strategic bet that AI will become essential to competitiveness, resilience and long-term performance in an increasingly complex energy system. “Long before disruption occurs, mission-critical industries must build resilience, while their increasingly complex, interconnected energy systems must deliver security, affordability and sustainable growth. Achieving this balance requires innovation that scales as operating conditions change. AI, combined with deep industry knowledge and operational data, can help operators make faster, more informed decisions — and provide intelligence critical to creating a smarter, more connected, and more resilient energy future.” Vimal Kapur Chairman and CEO, Honeywell • STRENGTHENED: The share of energy leaders naming AI as the highest corporate priority rose from 9% to 26%, while those placing it among the top two rose from 42% to 63%. • UNCHANGED: Nearly nine in ten organizations increased AI and digital infrastructure spending for a second year. • NEW: Almost two-thirds say existing investments will cover or exceed their costs. The 2026 perspective Powering Possible 2026 17

2 INTELLIGENCE MOVES INTO OPERATIONS AI is proving its value most where data is already well structured. The next test will be in production environments, where agentic systems must navigate greater uncertainty, turn insights into actions and deliver value closer to the asset where decisions are more consequential. Most energy organizations are past Looking specifically at agentic AI generation and production and the experimentation stage with AI. workflows, the findings are even 7% for grid operations. Only 1% report no AI use at all, and stronger, with agentic AI use more From the data, it’s reasonable to 16% are still piloting isolated use than doubling year-on-year, from assume that agentic workflows have cases. Half now run AI across several 15% to 33%. When industry scaled first in functions where data is | business functions, 26% have | respondents identify the functions in | | already structured, systems already | | integrated agentic workflows, and 7% | which agentic workflows are used | | connected and outcomes quickly | | orchestrate AI across organizational | most, digital and technology functions | | measured. Those environments allow | | or sectoral boundaries. | lead at 55%, compared with 8% for | | Figure 5 | | Over a third of businesses are starting to | | engage with agentic implementation of AI | | Organizational use of AI | | 1% | 4% | | 15% | 12% | 12% | 18% | 21% | 14% | 18% | 15% 51% 46% 45% 48% 52% 58% 42% 56% | 26% | 26% | 36% | | 26% | 19% | 22% | 38% | 24% | | 16% | | 7% | 8% | 8% | 7% | 6% | 2% | 2% | 2026 overall UK Germany Japan China US Brazil India Orchestrating Integrated, Several business Piloting isolated Not using yet across boundaries agentic workflows functions use cases 18

an organization to establish governance, prove reliability and build institutional confidence before extending the same capability into production environments, where data is more complex, equipment more varied and decisions carry physical consequences. For energy leaders, this raises a practical question: how can agents contribute more directly to the work of operating assets? “The question is no longer what AI can do, but how we deploy it effectively in our operations. In Subsurface, by combining decades of geoscience knowledge and know-how, petaflops of data with next- generation AI models, we can support our geosciences experts in the exploration, characterization and development of oil and gas reservoirs. In industry, performance goes hand in hand with reliability, safety and clear accountability. AI can help us operate our assets better and make better decisions. But technology must remain at the service of people. The real value will come from combining the power of AI with human expertise. It is a matter of augmenting the people capacities at their service.” Patrick Pouyanné CEO, TotalEnergies The opportunity is to connect insight with execution. Identifying a potential equipment fault is one task. Preparing an inspection plan, assembling the relevant records and coordinating an approved response can extend that capability. The ambition is moving beyond simply another recommendation and towards automated actions, always with human oversight. Autonomous operations require explicit choices about authority. Policies must define what an agent may read, recommend, prepare and execute—and where a person must approve the next step. Access to asset information for example should not automatically confer permission to change equipment settings. Checks have to match the consequences of an action, with an accountable operational owner and a clear fallback when information is incomplete or conditions change. This is both a leadership priority as well as a technical one. Seventy percent of leaders say AI-integration work took up a substantial amount of their professional time over the past twelve months.Sixty-eight percent now spend a quarter or more of their working day on tasks substantially involving AI.Fifty-six percent say AI influenced a significant number of their professional decisions in the same period.Intelligence has reached the executive layer of organizations; extending it to the operating asset at scale is the work now in front of the sector. Ultimately, the end goal for greater use of agentic AI in energy should focus on expanded capability, not autonomy for its own sake. Each and every workflow should answer a practical question: what can the organization do better because intelligence can act within an agreed scope? For energy businesses, the ambition is faster coordination, better-prepared responses and greater continuity of operations. How the market reads the sector Market expectations are running ahead of deployments. Investors describe a more automated industry than operators report. On grid operations, investors place agentic deployment at 20%, against the 7% operators report for their own organizations. On generation and production, investors say 13%, against 8%.While the two measures are not directly comparable — investors are assessing the sector, operators are reporting their own operations, and the bases differ — it is a gap that the sector has both the opportunity and the evidence to close. • UNCHANGED: 49% of organizations ran AI across several business functions in 2025, 50% today. • NARROWED: The pilot stage. A third of organizations were piloting isolated use cases a year ago; today it is one in six. That ground has gone to integrated agentic workflows, which rose from 15% to 33%. • NEW: Where the agents actually run. Digital and technology functions lead at 55%, while generation and production stands at 8% and grid operations at 7%. The 2026 perspective Powering Possible 2026 19

Energy leaders cite the same two primary barriers to adoption as 2025: Cybersecurity risks first, named by 43% of leaders, and data quality and consistency second, at 39%. Lack of skilled talent overtook cost to move from fourth to third, at 33%. For a technology moving as quickly as AI, the consistency is telling. Leaders are not discovering new problems with The industry's biggest AI challenges are the same as they were a year ago. As attention shifts toward agentic systems and greater autonomy, trusted data, robust security and clear governance remain the primary blockers to adoption. AI in energy. They are living with the same issues and describing them more precisely than they did in 2025. What has changed is what else appears on the list. Geopolitical risk enters at 24%. Limited cross-border AI coordination at 18%. Supply-chain risk at 10%. None of these made the list for last year’s report. The challenges have widened from the ones a company can fix itself to the ones it cannot — and 77% say AI's potential in the energy sector is held back because the system is too fragmented and siloed to apply it at scale. You can make a strong argument that cybersecurity and data quality are different vectors for a single underlying condition, namely whether an organization can rely on what its 3 CYBERSECURITY AND DATA READINESS REMAIN CHALLENGES FOR SCALE 20

“The energy sector has a good understanding of the technical challenges of adopting and scaling AI. The 2026 What we're seeing now are data says, and control who is able to challenges around see it, access and manipulate it. This perspective coordination, trust and perhaps explains why the priorities readiness at system level. leaders set for data integration are Organizations need not the ones a technical reading might • UNCHANGED: confidence in the data predict. Asked what should come The top three challenges, in they're using, clarity first, they name security, privacy and the same order: cybersecurity around accountability and regulatory compliance at 51% — risks, data quality and the ability to collaborate ahead of accuracy and reliability at consistency, across organizational 46%, and interoperability and lack of skilled talent. boundaries. Energy has common standards at 43%. • NARROWED: always been a connected That ordering is also reflected in the Loss of human oversight, industry. As AI becomes findings around data readiness for AI. a top-six concern in 2025 at more deeply embedded in Just eight percent of organizations 27%, is named by 21% this the system, the ability to say most of their data is structured year. Reputational risk, already work across those and AI-ready. A third say much of it is. the smallest concern at 9%, connections will become Four in ten say some of it is, and now stands at 7%. an even greater measure of its value.” 17% say little or none. • NEW: What seems clear from the above is Geopolitical risk Olivier Oullier that data demands from leadership (24%), limited cross-border CEO and Co-founder, Inclusive | are not for better pipes but for | AI coordination (18%) | Brains | | improved conditions under which | and supply-chain risk | | the pipes can be opened — clear | (10%) were not offered as | | ownership, defensible governance, | options in 2025. | | and permission to move data across | | a boundary and those conditions are | | set between organizations rather | | than inside them. | | Figure 6 | | Leaders cite and data quality issues as the | | key barriers to AI adoption | | Biggest barriers to AI implementation | | 41% 43% | | 39% | 2025 | 2026 | | 37% | | 32% 32% | 32% | | 29% | 30% | | 28% | 28% | 27% | | 24% | | 21% | | 18% | | 10% | 9% | 7% Cybersecurity Data Lack of Unpredictable/ Regulatory Upfront Geopolitical Loss of Limited Supply Reputational risks quality and skilled escalating AI risk cost of risk human cross-border chain risk risk consistency talent usage costs implementation oversight AI coordination Powering Possible 2026 21

Leaders are nearly unanimous when asked about the near-term energy demands of AI. Seventy-four percent expect that greater adoption of AI will increase the total amount of energy their country consumes by 2030. Six percent expect a decrease. The IEA puts global data-center electricity demand at roughly 950 TWh by 20304 (IEA). The longer view is less settled, and potentially more interesting. By 2050, 64% still expect AI to raise total energy use — but 24% now expect it to lower it, roughly four times the number who say so for 2030. The data suggests that the efficiency case is both established and that the industry believes in it. Asked where AI will most affect the energy transition, 57% name optimizing complex systems — power flow, dispatch, balancing — well ahead of automating processes at 39% and unlocking scientific discoveries at 32%.The expected payoff is therefore, in the near term at least, not a new technology, it is a better-run system. This potentially presents a challenge when allocating capital: demand rises first, and the intelligence that moderates it pays back later. Twenty- four percent of leaders and 23% of private equity and venture capital name sovereign wealth and state- directed investment as the most influential sources of capital by 2030, with institutional investors at 15%. By 2050 venture capital falls to 17% and institutional money rises to 18%, while sovereign and state-directed capital dips slightly to 23% — the only source Most leaders expect AI adoption to increase national energy use by 2030. The investment task is to build capacity and intelligence together — coordinating generation, networks and digital infrastructure while pursuing efficiency gains. expected to matter as much in twenty-five years as in five. Private capital leads where returns are nearer term, while sovereign and state-directed investment remains important where returns take longer to materialize. The scale of corporate commitment is rising alongside it. BCG’s cross- industry AI Radar report cites executives in energy and utilities projecting AI investment climbing from 0.6% to 1.9% of revenue5. But the binding constraint is downstream of any model. The IEA puts annual investment in electricity grids at roughly $400 billion — well under half of what flows into generation6 (IEA). Capacity is being built faster than the network that has to carry it. That constraint extends upstream of the power plant. Expanding gas supply and LNG capacity does not by itself create reliable power. Gas must be processed, transported, stored and delivered to generators, while networks must carry electricity to concentrated demand. Because these assets operate on different investment cycles and are unevenly distributed, an underdeveloped link can become the bottleneck. Supply, transportation, storage, generation and grids should therefore be financed as a connected system. Our data points at the same gap. Asked for the single most important policy action to support sustainable AI and energy growth, leaders name grid investment first at 24%, with AI and energy regulation level at 24% and incentives for clean energy at 17%. Two options were new this year — energy system integration, and cross-border coordination frameworks — and together they took 29%. From the findings you can make the case that both the money and policy are being drawn toward the same place: not technology, but the system it runs on. 4 CAPITAL AND POLICY ARE BEING DRAWN TOWARDS THE GRID • UNCHANGED: Near-term expectation. Seven in ten senior leaders expected AI to increase national energy consumption by 2030 a year ago, and three-quarters do now. • NARROWED: The share expecting AI to reduce energy use by 2030 has roughly halved, from 12% to 6%. At 2050 it has fallen too, from 30% to 24%. Leaders have grown more convinced that AI adds load before it saves any. • NEW: Where the capital comes from. Private capital dominates the near horizon and institutional money the far one, but sovereign wealth and state-directed investment holds at 24% across both. The 2026 perspective 4 Key Questions on Energy and AI, World Energy Outlook Special Report, IEA, April 2026 5 BCG AI Radar 2026: As AI Investments Surge, CEOs Take the Lead, Boston Consulting Group, January 2026 6 World Energy Investment 2025, IEA, June 2025 22

Figure 7 Sovereign wealth and private equity are seen by leaders as the main sources of investment for the future energy system Most Influential Source of Capital in Investing in the Future Global Energy System | 24% 23% | 23% | 2030 | 2050 | | 17% | 18% | 14% 11% 11% 12% 10% 9% 7% 6% 6% Sovereign wealth Private equity and Institutional Commercial banks Multilateral Corporate balance Bond markets funds and state- venture capital investors (pension and project development sheets and (including green directed investment funds, insurance finance debt finance reinvested profits and climate bonds) companies, asset managers) “ We're entering a period where demand is arriving faster than infrastructure can be built. AI is adding to that challenge, increasing the need for power today while much of its economic value will emerge over time. That places a premium on long-term thinking. Investors are looking beyond individual assets and asking whether the broader system can support growth reliably and at scale. The opportunity is significant, but capturing it will depend on coordinated investment across generation, networks, storage and digital infrastructure.” Ken Moelis Founder and Executive Chairman, Moelis & Company Powering Possible 2026 23

To date, most conversations around energy and AI describe software. Reflecting the evolution of the technology, this year’s report asked about something else: physical intelligence — robots, drones, smart equipment — whose actions and decisions are driven by AI rather than by pre-programmed automation. The answers describe a familiar shape at an earlier stage. Sixty percent of industry respondents are now using autonomous or robotic systems in physical operations against 36% yet to deploy. Across all industry respondents, 30% are piloting them in isolated, single-task settings, 24%, use them routinely for defined tasks, and a small but significant 6% run them as a connected layer, where AI decision-making leads to autonomous actions in the field. Set that against software, where less than 1% of industry respondents are not using AI at all and just14% are still piloting.These findings suggest physical intelligence is at an earlier stage of adoption than software AI. Finance and investment leaders see it AI is beginning to leave the control room. New for 2026, leaders report autonomous and robotic systems moving into physical operations. The opportunity is safer operations, higher-value work for skilled teams and a direct link between insights and action in the field. differently, as they do on agentic workflows. When asked for their view of the sector, only 13% think energy companies have not yet begun, against the 36% of industry respondents reporting exactly that; more than half believe the sector is piloting.The two questions are asked of different groups, with different framing and different bases — but this is the second time in this section that the market reads the industry as further along than the industry reports. One such gap can be seen as a curiosity, two suggests a pattern. What leaders want from these systems is revealing. Asked to rank where autonomous and robotic systems will deliver the greatest strategic value over the next three years, two answers separate from the rest: efficiency — reducing routine manual workload and physical effort — selected by 44%, and predictive action — maintaining and protecting critical assets before they fail — selected by 41%. Workforce augmentation and safety and access follow, each on about a third.The case leaders make for putting intelligence into the field is strongly practical and protective. It is about keeping assets running and people out of hazardous places, not about removing them. Looking beyond the report's findings, the World Economic Forum and Boston Consulting Group's Physical AI: Powering the New Age of Industrial Operations whitepaper sets out three systems that will coexist across industry: rule-based robotics for structured, repetitive work; training-based systems that learn variable tasks; and context-based machines that interpret unfamiliar surroundings and act7. That forecast explains what the survey shows, with one caveat; rule-based automation rewards structure and repetition, and energy's physical environments — offshore platforms, remote pipelines, live process plants — offer little of either. Energy has less structured work to automate, so it is entering physical AI closer to the difficult end — which makes the sector's 60% a more substantial figure than perhaps it first appears. 5 INTELLIGENCE IS GETTING PHYSICAL 7 Physical AI: Powering the New Age of Industrial Operations, World Economic Forum in collaboration with Boston Consulting Group, September 2025, 24

Seventy-seven percent of organizations increased their spending on upskilling workers to use AI over the past twelve months, and 35% describe that increase as significant.Asked what matters most to incorporating more AI into the energy sector, 29% of leaders name upskilling energy-sector professionals, almost exactly the share who said so a year ago8. This is a constraint the sector identified early, has held to, and is now putting real money against. Upskilling investment is rising in every market surveyed, ranging from 88% in India to 63% in Brazil. However, the work is not finished. Lack of skilled talent remains the third-largest barrier to adopting AI in the sector, named by 33% of leaders, behind only cybersecurity and data quality.Both of those figures are worth holding together: an industry that is spending more on skills and still ranking them third suggests something about the size of the task, and the gap that still needs to close. If the industry's challenge is turning AI skills into AI capability, ADNOC shows what that looks like in practice. The most significant measure of ADNOC's AI progress is perhaps not how many people have been trained – although that is considerable – but how many are putting AI to work. Across AI may automate more work, but it will not automate accountability. As intelligent systems become more capable, the role of people increasingly shifts towards judgment, oversight and decision-making. The priority for leaders is developing the skills needed to work effectively alongside increasingly autonomous systems. operations, engineering, commercial and corporate functions, employees are using AI to solve problems, streamline processes and make better decisions. Thousands of AI agents have been built and deployed by employees themselves, each designed around a real business need. Many of these solutions come not from dedicated technology teams, but from the people who understand the challenge best: front-line engineers, operators, analysts and business professionals. ADNOC is demonstrating what happens when AI moves beyond experimentation and becomes a capability embedded across the workforce, turning everyday expertise into scalable digital solutions. The broader industry appears to be moving in the same direction. Fifty- two percent of leaders now spend six hours or more a week learning about AI or digital topics, up from 29% a year ago, while the share committing eleven hours or more has tripled to 25%. Several hundred senior executives have independently decided to devote something close to a working day each week to building their own understanding of the technology. Capability, it seems, is increasingly being treated as a leadership responsibility rather than a training program. That helps explain another finding. When leaders look for external support, skills programs rank surprisingly low on the list. Presented with six possible actions to support sustainable AI and energy growth, they put grid investment and AI regulation first, followed by clean-energy incentives, system integration and cross-border coordination frameworks. Digital skills programs come last, attracting just 6% of responses, down from 8% a year ago. The implication is not that skills matter less, rather, leaders increasingly see capability as something that must be built inside their organizations, while governments focus on the infrastructure, regulation and coordination needed to support AI at scale. Looking across the findings for Powering Possible 2026, agentic workflows have reached the functions whose data was ready. Data readiness is turning into a question of governance and permission. Intelligence has begun to act on the physical asset. Every one of those findings resolves to people: a person or a team capable of making the decision, and accountable for having made it. The systems described across these pages do not run themselves, and the organizations getting the most from them are the ones whose people are learning, including their leaders. 6 PROGRESS DEPENDS ON PEOPLE 8, 9 Unleashing AI for Energy and Energy for AI, ADNOC and Microsoft, October 2025 26

Figure 9 Leaders see upskilling and cybersecurity as key to integrating AI in the energy sector Most important improvement to integrate AI in the energy sector | 31% | 29% | 31% | 2025 | 2026 | | 27% | 27% | | 23% 25% | 24% | 23% | | 21% | 22% 20% | | 17% 18% | 18% | | 15% | | 7% | 7% | Improving Upskilling of Advanced Sensor Data Standardizing Data Improving Improving cybersecurity energy sector AI model deployment infrastructure data formats integration regulatory communications frameworks professionals development upgrades & protocols upgrades framework networks The 2026 “Efficiency is the foundation of every AI business case at perspective ADNOC, and it's where we start. But it isn't where the value stops. When AI helps • UNCHANGED: a facility keep running Upskilling energy-sector through disruption or keeps professionals is named a team out of a hazardous among the most important space, that's a return too. enablers of greater AI use by The next phase of AI in 29% of leaders, against 27% energy belongs to those a year ago. who can measure the full picture, not just the part • NARROWED: that's easiest to count.” A third of leaders spent one to two hours a week Nada Al Meqbaali learning about AI last year. SVP Strategic Tech & Over the same period the AI Partnerships, ADNOC share committing eleven hours or more has tripled, from 8% to 25%. • NEW: Leaders were asked for the first time this year how their spending on upskilling workers to use AI has changed. More than three-quarters report an increase, and a third describe it as significant. Powering Possible 2026 27

The leading measure for AI's role in the energy transition may not have changed since 2025, but confidence has soared. Energy supply stability and reliability, the leading measure for two years running, is named by 39% of leaders, up 13 percentage points from 26% in 2025. Prediction accuracy and data use has fallen from 24% to 19%. Cost savings also declined from 19% to 13%. These changes suggest that leaders are looking beyond the performance of an individual model and towards the performance of the system around it. A prediction can be accurate and still have little value if it does not lead to a better decision, a safer operation or a more dependable supply of energy. Energy supply stability and reliability are the now top measures leaders use to judge AI’s success. As token-based consumption turns it into an ongoing operating cost, the business case must show what AI delivers for continuity and resilience, not only productivity and savings. What matters is whether the system continues to work when conditions become difficult. They expect the greatest contribution in areas that depend on coordination. Fifty-seven percent say AI will have its biggest impact through the optimization of complex systems, including power flows, dispatch and the balancing of demand across a network. Process automation follows at 39%, and scientific discovery at 32%. The common thread is the ability to keep many moving parts aligned, continuously and at a scale that would be difficult for people to manage alone. Leaders are confident about the direction of travel, although less certain about the pace. Fifty-four percent expect AI to support and improve energy resilience by 2030. By 2050, that rises to 82%. Because both questions were asked at the same time, the gap is not a measure of changing opinion. It is a judgement about how quickly the benefits will emerge. There is a significant minority that remains unconvinced. Thirteen percent expect greater use of AI to decrease energy resilience by 2030, while 31% expect it to make no difference at all.Confidence is strongest in India, where 78% believe AI can make energy supply more 7 RESILIENCE IS EMERGING AS A MAJOR DRIVER OF AI INVESTMENT AI expected to boost energy resilience by majority of leaders Figure 10 2030 2% 2050 3% Support and improve energy resilience Have no impact on grid and energy resilience Decrease energy resilience Don’t Know Impact of the greater use of AI on energy resilience 54% 31% 13% 82% 10% 5% 28

resilient, compared with 49% in Germany, underlining that the direction of travel is shared more widely than the level of conviction. The broader conviction, however, has held steady. Three-quarters of leaders believe AI will improve energy production by increasing safety, reducing emissions and improving efficiency, almost exactly the same proportion as last year.That view has not shifted despite a much louder debate about the cost of AI and the resources required to run it. Those costs are becoming more complicated to understand. The price of an individual token may fall, but production systems can consume tokens at scale through repeated model calls, longer exchanges and increasingly autonomous workflows. Total spending can therefore rise even as the unit price falls. This makes the question of value more, not less, important: which uses deliver a measurable improvement in reliability, safety or resilience, and which simply add another layer of consumption? That also brings accountability into sharper focus. If reliability is becoming the test of AI’s contribution, someone must be answerable for the decisions that shape it. As more decisions are automated, coordinated across assets and taken across borders, the question is no longer only whether a system can act. It is who is responsible for the outcome. • UNCHANGED: Seventy-six percent of leaders said that AI would optimize energy production through safety, emissions and efficiency, virtually the same as last year. • NARROWED: A fifth of leaders named cost savings as the most important metric for judging AI in the energy transition last year; one in eight does now. • NEW: AI value is becoming a unit-economics question. As token prices fall and usage grows, AI must prove the value of each workflow and decision. The 2026 perspective “As AI accelerates across the global energy sector, it risks widening Africa's energy poverty and investment gap. While advanced economies use AI to make energy systems more efficient, competitive and profitable, African projects could face higher costs, greater competition for capital, skills and infrastructure, and increased barriers to financing. Yet this outcome is not inevitable. By strengthening investment frameworks, developing bankable projects, investing in human capital and using regional institutions to drive collaboration and common standards, Africa can secure its place in the emerging AI-driven energy economy. The opportunity is significant, but so is the urgency.” Proscovia Nabbanja Uganda National Oil Company (UNOC) Powering Possible 2026 29

Asked what the top priority for governing AI should be, as it increasingly coordinates energy across borders, energy leaders answer in the language of specification. Security and system reliability comes first, at 53%. Transparency and accountability of AI decisions follows at 44%, and common technical standards and interoperability at 40%. Protecting national energy sovereignty is named by 24%.46 Read as a design brief, it is extremely coherent: make it secure, make its decisions inspectable, and make its parts work with one another. These are requirements rather than candidates — a description of the properties leaders want the system to have, whoever builds it. They are also clear about where to start. Asked which areas of the energy sector would benefit most from sector-wide data standards for AI, 41% name grid management and optimization — more than twice the next answer, predictive maintenance at 18%.47 A year ago the same question produced the same first answer at 39%.48 Through a period in which almost everything about AI in energy moved, this one conviction has not. It is the closest thing the report contains to a settled engineering priority. As AI begins coordinating energy systems across assets, companies and borders, leaders are converging on a clear set of priorities: security, transparency and shared standards. The priority now is creating the conditions for AI to operate safely, reliably and at scale. The supporting picture points the same way. Leaders were asked what is most important to incorporating more AI into the energy sector, and the answers have tilted over 12 months toward the architecture around the technology rather than the instruments feeding it: improving cybersecurity frameworks is named by 31%, up from 27%, and improving the regulatory framework by 18%, up from 15%, while sensor deployment has fallen from 31% to 23%.49 The sector spent recent years working out how to measure things. It is now working out how to govern what the measurements set in motion. One item on that brief is missing, however. Forty-four percent of leaders want AI's decisions to be transparent and accountable. Far fewer — 18%, the second lowest- ranked of the six priorities — ask for clear lines of liability when a system fails.50 The difference between the two is important. Transparency tells you why an automated decision was made. Liability settles who is answerable for it. As AI takes on decisions that move energy between markets, the second question becomes harder to leave open. That gap aside, what stands out is that this year’s report lands in familiar territory. When leaders described what holds AI back, the obstacles were about permission and ownership more than technology. When they were asked what policy should do, they reached for system integration and cross-border coordination. Now, asked how AI should be governed, they specify standards, interoperability and shared conditions. Three different questions, one consistent answer: the work that matters most sits between organizations, not inside them. What leaders are describing is infrastructure as much as policy. Standards, interoperability and the ability to verify a decision must be built, run and powered. The objective should be dependable cooperation with clear control, not technological self-sufficiency. 8 COOPERATION IS KEY FOR AI GOVERNANCE 30

“The energy sector has always depended on cooperation. Energy moves across assets, companies and national borders through systems that only function when participants agree on standards, responsibilities and ways of working. As AI becomes more deeply embedded in those systems, the same principle applies. The objective should not be uniformity, but trust: confidence that systems can work together securely, reliably and transparently, even when ownership and interests differ.” Takayuki Ueda President & CEO, INPEX Security, transparency and shared standards lead priorities for cross-border governance according to energy leaders Figure 11 53% Security and system reliability 44% Transparency and accountability of AI decisions 40% Common technical standards and interoperability 24% Protecting national energy sovereignty 18% Clear lines of liability when systems fail 15% Fair and equitable access across regions Top priorities for governing AI coordination across borders • UNCHANGED: Grid management and optimization was named by 39% of leaders a year ago as the area that would benefit most from sector-wide data standards for AI, and 41% name it now. • NARROWED: Sensor deployment was named among the most important enablers of greater AI use by 31% of leaders last year; 23% name it now. Attention has moved from gathering more data to governing what the data sets in motion. • NEW: What leaders want governing AI across borders. Security and system reliability leads at 53%, ahead of transparency and accountability of AI decisions at 44% and common technical standards at 40%. The 2026 perspective Powering Possible 2026 31

The energy sector's AI ambitions increasingly converge on a single piece of infrastructure: the grid. It is where rising electricity demand becomes a physical reality, where investment decisions play out, and where leaders increasingly expect AI to prove its worth. Yet one of the central questions facing the industry remains unresolved. Fifty-two percent of energy leaders are confident existing infrastructure can meet AI-driven electricity demand through 2030. Forty-seven percent are not. The industry was divided on the question in 2025 and remains divided today. The global split conceals a sharper geographic divide. Confidence reaches 75% in Brazil and 70% in India, but falls to 40% in the US and 38% in the UK. What has changed is the scale of what is being discussed. The IEA projects data-centre electricity demand will reach roughly 950 TWh by 203010, increasing the importance of a question leaders have yet to settle: can energy infrastructure keep pace with growing demand for intelligence? Leaders are placing greater confidence in AI’s ability to strengthen grid capacity and reliability, but no more confidence in the infrastructure needed to support it. The resulting gap between expectation and readiness is becoming one of the defining challenges of AI’s expansion. Leaders may be divided on whether existing infrastructure can keep pace, but there is far more agreement on what should happen next. Asked which investments would most improve grid capacity and resilience to support AI growth, 68% name grid modernization and smart grid technologies, placing them first by a wide margin, as they did a year ago. Energy storage follows at 42%, with real-time monitoring at 29%. The pattern repeats elsewhere. Asked how to get more from existing generation, leaders again put grid flexibility first at 44%, followed by accelerating the connection of new projects at 32%. Two years of data point to the same conclusion: while uncertainty remains over future demand, there is growing consensus on where investment should be directed. One thing has moved. Fifty-nine percent now expect AI to have a large impact on increasing the capacity and reliability of the grid, against something closer to 45% a year ago . Leaders have gone from expecting AI to help, to expecting it to help substantially. That expectation now sits at the center of the sector's thinking. The grid is where leaders said shared data standards should begin, and where they directed their first policy ask. It is also, on this evidence, where they expect AI to earn its keep. When the first Powering Possible report was published in 2024, energy for AI and AI for energy were emerging as two sides of the same conversation. Two years on, that relationship is becoming increasingly difficult to separate. The intelligence described throughout this report depends on energy systems that are reliable, resilient and able to grow with demand. At the same time, leaders increasingly expect AI to help build exactly those capabilities. The future of both will be shaped together. 9 GRID MODERNIZATION REMAINS KEY CONCERN 10 Key Questions on Energy and AI, World Energy Outlook Special Report, IEA, April 2026 32

“When grounded in high- The 2026 quality data, AI can amplify human expertise, improve perspective decision-making, and help businesses tackle increasingly complex • UNCHANGED: challenges. Creating long- Around half of leaders term value means moving were confident existing beyond individual AI infrastructure could meet | applications and rethinking | AI-driven electricity demand | | how our work gets done.” | through 2030 a year ago, and | | 52% are now. | | Wael Sawan | | CEO, Shell | • | NARROWED: | | Accelerating the connection | | of new projects drew close to | | two in five leaders last year | | as the most effective way | | to raise output from existing | | generation; 32% name it now, | | while enhancing grid flexibility | | has pulled clear at 44%. | | • | NEW: | | Fifty-nine percent now expect | | AI to have a large impact on | | grid capacity and reliability, | | against something closer to | | 45% a year ago — | | a hardening of a view already | | held by nine in ten. | | Figure 12 | | Regional differences appear among | | leaders on the grid's ability to meet future | | energy demands | | Confidence in existing infrastructure meeting grid demands | | 1% | 2% | 4% | 1% | | 25% | 30% | | 36% | | 47% | 54% | 51% | 62% | 59% | | 52% | | 52% | 45% | | 38% | 26% | 37% | | 28% | 32% | | 14% | 23% | 18% | 17% | 16% | 12% | 10% | 8% | 2026 overall Brazil India China Japan Germany UK US Very confident Somewhat confident Not confident Don’t know Powering Possible 2026 33

KEY NSIGHTS AND RECOMMENDATIONS

Energy Resilience in the Intelligence Era The findings for this year’s Powering Possible report describe a sector that has settled the question of whether AI belongs in energy and moved on to harder ground: where to deploy it, what to measure, and who is accountable when it acts. The task ahead is to build the systems that intelligence can run on, and to value what those systems protect. INSIGHTS RECOMMENDATION The measure of AI’s value has widened. Leaders are increasingly judging AI on its ability to help organizations navigate complexity, uncertainty and disruption, not just drive efficiency and reduce costs. Invest in AI as resilience infrastructure. As energy systems become more exposed to disruption, AI should be viewed as a resilience capability, not just a productivity tool. Predictive analytics, intelligent monitoring and optimization can help anticipate shocks, adapt faster and maintain performance under changing conditions. Investment momentum is clear; conversion is the test. One in five leaders say current investments in AI have clearly exceeded their costs, while more than half expect future investments to do so. Confidence still runs ahead of proof. Close the gap with evidence. Publish use cases and realized outcomes from completed deployments, and hold new investments to clear KPIs for operational, financial and resilience value. 36

INSIGHTS RECOMMENDATION Agentic AI is moving from experimentation to enterprise deployment. Adoption has more than doubled since 2025, but remains concentrated in digital and technology functions. The next opportunity is extending agents into operational environments as governance, oversight and trust mature. Build for scale, not just deployment. Establish trusted data foundations, governance and operating models that allow agents to be deployed consistently across business and operational environments. The top two barriers are one problem. Cybersecurity and data quality lead the list for a second year. Only 8% of organizations describe most of their data as AI-ready, and leaders prioritizing data integration put security and compliance ahead of accuracy. Run data readiness as a governance program, not an IT upgrade. Name owners for the data agents will act on, define who may move it across which boundaries, and resolve permission before pipelines. Capital and policy both point at the network. Grid modernization is the leading infrastructure priority, and grid investment is leaders’ first policy ask. Annual global investment in grids remains far below what flows into generation11. Invest for system resilience, not single-point optimization. Meeting future demand will require balanced investment in generation, networks, storage and digital infrastructure to improve capacity, reliability and flexibility across the energy system. Physical intelligence is arriving, and its value is protective. Thirty percent of operators are piloting autonomous and robotic systems and 24% use them routinely. Leaders expect the greatest value in efficiency and predictive action: protecting critical assets before they fail and keeping people away from harm. Start where it removes risk. Deploy physical AI first in hazardous, remote and hard-to-reach work, and manage it as an extension of asset integrity and safety as much as a robotics program. The sector is funding its own capability. Seventy-seven percent of organizations increased upskilling spend, leaders are spending more time learning about AI, and skills programs rank last among the policy actions leaders ask for. Invest deeply in in-house capability and make it measurable. Agree expected levels of AI fluency by function, track them as a safety competency, and train people to work across energy and intelligence systems. Accountability is agreed; liability is not. Forty-four percent of leaders want AI decisions to be transparent and accountable. Only 18% ask for clear lines of liability when a system fails, the second-lowest of six governance priorities. Settle liability before autonomy scales. Write responsibility for automated decisions into contracts, partnership agreements and operating procedures before AI decisions become more consequential. 11 World Energy Investment 2025, IEA, June 2025 Powering Possible 2026 37

Methodology Quantitative survey: recruited by individuals Total | • High knowledge, high seniority stakeholders within the | Audience | Interviews | | finance and energy sectors. | | • This audience was bespoke recruited for this study. | All | 398 | | • 398 quantitative interviews took place with this audience | Brazil | 49 | | The margin of error for the total quantitative sample is | China | 50 | | +/- 4.9% and median length of interview (LOI) was | | 25 minutes. | Germany | 50 | | India | 50 | | Japan | 49 | | UK | 50 | | US | 100 | 38

Powering Possible 2025 to 2026 Twelve months on, AI has moved up the corporate agenda and agentic workflows are taking the place of pilots. Leaders are starting to measure AI by the reliability it brings, and grid modernization is still their first investment choice. Change | Measure | 2025 | 2026 | (pts) | What it tells us | | Where the year moved | | AI as organization's top priority | 9% | 26% | +17 | AI has returned to the top of the | | corporate agenda | | AI is the top, or one of the top | 42% | 62% | +20 | Almost two-thirds now rank AI in | | two, priorities | their top two | | Organizations running integrated | 15% | 33% | +18 | Agents are moving into everyday work. | | agentic workflows | | Organizations piloting isolated | 35% | 16% | −19 | Fewer than half as many organizations | AI use cases remain at the pilot stage Supply stability and reliability is the 26% 39% +13 Reliability has pulled clear as the test of leading measure of AI's success AI's value Expect AI to have a large impact on 45% 59% +14 Leaders now expect AI to make a substantial grid capacity and reliability difference to the grid Expect AI to reduce national energy 12% 6% −6 Most now expect AI to add load before it use by 2030 saves energy Sensor deployment is the most 31% 23% −8 Attention is moving from gathering data important enabler of AI to governing it What barely moved Organizations that increased 85% 89% +4 Investment momentum continues AI investment for a second year Confident existing infrastructure 52% 52% 0 The sector remains split on can meet AI-driven demand to 2030 grid readiness Grid modernization is the top 64% 68% +4 The first choice again, by a investment to support AI growth wide margin Cybersecurity is the leading barrier 41% 43% +2 Remains the barrier leaders to AI adoption name most Grid management would gain most 39% 41% +2 Still the primary beneficiary for most from shared data standards energy leaders Powering Possible 2026 39

#ENERGYFORLIFE