Two applications go first

Field master planning comes first. Subsurface, facilities and economic studies currently run one after another. The new system merges them into one decision. OpenAI’s reasoning models will process decades of data from thousands of wells, multiple stacked reservoirs and hundreds of surface facilities, alongside capital and commercial factors. ADNOC expects to prepare complex development plans in hours rather than months. The plan optimises the whole field at once, not each reservoir or facility on its own.

The second application values every barrel while it is still in the ground. ADNOC says no other energy company uses the technology this way. Interconnected agents will forecast crude supply, demand, prices and margins continuously. They will weigh the commercial and operational impact of each decision across production, processing, refining, logistics, sales and trading.

Help with better decision making

Commenting on the deal Dr. Sultan Ahmed Al Jaber, UAE Minister of Industry and Advanced Technology and ADNOC Managing Director and Group CEO, said: “The next frontier of AI is not simply doing things faster. It is making possible what was previously beyond reach.”

Sam Altman, CEO of OpenAI, calls it a “first-of-its-kind partnership” that will help ADNOC “move faster, operate more efficiently and make better decisions across the energy value chain.” 

Mansoor Al Mansoori, CEO of G42 International, ties the deal to national strategy. “The UAE is building an AI native nation with a global purpose to export intelligence responsibly as we have long exported energy,” he says. 

The agreement builds on ADNOC’s rollout of ChatGPT Enterprise to more than 40,000 employees. That is the largest enterprise deployment of ChatGPT in the Middle East and North Africa, with adoption above 97 per cent.

It also complements ENERGYai, the subsurface AI ADNOC developed with AIQ. The tool runs on a 70-billion-parameter model trained on more than 50 years of company data. ADNOC operates more than 200 AI applications and 70 robotics use cases, deployed or under development. Its Real-Time Operations Center connects more than 120 rigs in one live environment and has cut engineering effort by 30 per cent to 40 per cent.