ADNOC Adopts OpenAI Platform Under Multiyear Partnership for Energy Applications
Key Facts
ADNOC, OpenAI and G42 announced a multiyear partnership on November 2 at the ADIPEC exhibition in Abu Dhabi to deploy the OpenAI platform and connect AI agents to the company’s operations and decisions across its value chain. ADNOC said it is the first energy company to deploy OpenAI’s advanced AI platform. The announcement comes as Sam Altman has described energy as AI’s biggest challenge, with demand continuing to rise. The partnership puts industrial applications of the technology at the center of OpenAI’s expansion in the UAE, without announcing operational results or realized savings so far. GPT models will also run on G42’s sovereign AI infrastructure in the UAE.
Sultan Al Jaber, ADNOC’s chairman, said the move aims to “connect intelligence across our entire business.” The partnership will begin with two applications the company described as high-value. The first focuses on integrated field development planning, while the second links operational and commercial information across the company’s business stages. These include production, processing, refining, logistics, sales and trading. The announced use thus extends from analyzing field data to assessing commercial and operational decisions across the value chain.
For the first application, data from wells, reservoirs and surface facilities will be brought together to develop integrated field plans. ADNOC says the planning will incorporate decades of data from thousands of wells, multilayer reservoirs and hundreds of surface facilities. The aim is to move from separate studies to optimizing the plan across the entire field. GPT models will run on G42’s sovereign infrastructure in the UAE, according to the statement. The company said human experts will retain oversight of decisions with significant consequences, a stated operational limit on the systems’ role.
The second application will use interconnected AI agents to forecast crude oil supply, demand, prices and profit margins. ADNOC said these forecasts will be used to estimate the effects of variables across the value chain before they occur. The announcement did not specify impact metrics or a timeline for measuring the applications’ results. The company says the partnership builds on the deployment of ChatGPT Enterprise among more than 40,000 employees, with adoption exceeding 97% when the agreement was announced. That existing rollout provides the backdrop for expansion from employee tools to applications directly tied to operations and corporate decisions.
At ADIPEC, Altman said energy is “the most dominant factor” in determining how much AI the world can use. He added that the energy required per query is falling as efficiency improves, but lower usage costs encourage greater demand. He said he hoped AI applications would deliver “a net gain for the world’s energy availability.” This places the ADNOC partnership in the context of a dual relationship: the technology consumes electricity and may also be used to support energy-sector operations. The International Energy Agency’s 2025 report estimated that global data center electricity consumption would reach about 945 terawatt-hours in 2030, with AI the most important driver of the expected growth.
International Energy Agency data for 2026 show accelerating demand linked to data centers: their electricity consumption rose 17% in 2025, compared with 3% growth in global electricity demand. The agency said data center consumption could double by 2030. It warned of pressure on supply chains for gas turbines, transformers, advanced chips and IT components, as well as delays in connecting centers to the grid and in approval processes. These constraints reflect a challenge that extends beyond model efficiency to the availability of the equipment, energy and infrastructure needed to run them. ADIPEC is scheduled to take place in Abu Dhabi from November 2 to 5, with discussions bringing together technology and energy leaders on integrating AI into infrastructure and operations.