OpenAI Signs Four Abu Dhabi Deals with ADNOC and Mubadala
OpenAI has signed four strategic agreements with organizations in Abu Dhabi, as CEO Sam Altman visited the UAE and met Crown Prince Sheikh Khaled bin Mohamed bin Zayed Al Nahyan during the opening of ADIPEC 2026.
The agreements span government services, energy, investment, and education. They include a new initiative to make government regulations readable by AI systems, a partnership with Abu Dhabi National Oil Company (ADNOC) to introduce OpenAI Frontier into energy operations, and an agreement with Mubadala Investment Company to apply AI to investment research and risk assessment. A separate five-year partnership will support the adoption of OpenAI technologies across UAE government entities, businesses, and universities.
The announcements coincided with Altman’s appearance at the opening of the Abu Dhabi International Petroleum Exhibition and Conference (ADIPEC), attended by Dr. Sultan Ahmed Al Jaber, UAE Minister of Industry and Advanced Technology and ADNOC’s Managing Director and Group CEO. During the event, Altman discussed the future of AI, the industry’s growing energy requirements, and the importance of keeping humans in control of critical decisions.
Abu Dhabi Launches a Machine-Readable Government Initiative
The Abu Dhabi government has partnered with OpenAI to launch its Machine-Readable Government initiative, an effort to prepare public-sector rules and procedures for use by AI systems.
The initiative converts selected government policies and regulations into structured digital formats that AI systems can interpret. This allows them to review applications, compare submitted information against official requirements, and produce assessments based on applicable rules.
Abu Dhabi’s Department of Government Enablement (DGE) is leading the initiative and overseeing its governance. The first practical application is being introduced at the Abu Dhabi Agriculture and Food Safety Authority, focusing on the inspection of imported food shipments.
The system analyzes shipment information alongside the relevant regulatory requirements, then produces assessments to help government employees determine whether shipments meet the applicable standards. Each assessment is linked to the specific regulation behind it, allowing officials to review the reasoning and verify its basis. Government employees retain oversight and responsibility for final decisions.
The initiative is expected to expand beyond food-import inspections. Planned applications include building permits, business and freelance licenses, data-access governance, human resources compliance, and government procurement.
Data processing will take place locally within the UAE using GPU infrastructure hosted in the country. Under the announced framework, the Abu Dhabi government retains full ownership of the data generated through the system, and that data will not be used to train AI models.
ADNOC Becomes the First Energy Company to Deploy OpenAI Frontier
In the energy sector, ADNOC has signed a strategic agreement with OpenAI and G42, through its subsidiary Inception42, to introduce advanced AI technologies into the company’s operations.
Under the partnership, ADNOC will become the first energy company to deploy OpenAI Frontier, a platform designed to help organizations build and operate AI agents connected to their internal data and business systems.
The initial focus is on developing integrated plans for oil field management. The work will draw on decades of operational information covering thousands of wells, oil reservoirs, hundreds of surface facilities, investment costs, and commercial considerations.
The goal is to produce and refine field development plans in hours rather than months by bringing geological, operational, and financial analysis into a unified workflow.
Another area of collaboration will connect decisions across production, processing, refining, transportation, sales, and trading. AI agents will analyze supply and demand forecasts, oil prices, and profit margins, while assessing the potential impact of operational and commercial decisions before they are implemented.
The partnership will use GPT models alongside sovereign computing infrastructure located in the UAE. Human specialists will continue to oversee important decisions.
ADNOC also plans to expand employee training and adoption of AI tools. According to the company, ChatGPT Enterprise has already been deployed to more than 40,000 employees, with usage exceeding 97%.
Mubadala to Apply AI to Investment Research and Risk Analysis
In the financial sector, Mubadala Investment Company has signed an agreement with OpenAI and Inception42 to explore the use of AI in investment analysis, opportunity identification, and risk assessment.
The collaboration will focus on analyzing investment data, researching markets and industries, and evaluating companies. These capabilities are intended to give investment teams broader analytical support when assessing opportunities and examining potential risks before making decisions.
The organizations did not disclose a financial value for the agreement. The announcement also did not include any direct investment by Mubadala in OpenAI.
Five-Year Partnership to Expand OpenAI Technology Across the UAE
The fourth agreement brings OpenAI and Inception42 together in a five-year partnership aimed at accelerating the adoption of advanced AI technologies across UAE government entities, businesses, and educational institutions.
The two companies will work on enterprise solutions for large organizations, combining OpenAI’s models and engineering expertise with Inception42’s capabilities in AI application development, consulting, and professional services.
The collaboration will cover government, business, and education, with an emphasis on integrating AI into organizational workflows and building applications suited to the needs of UAE institutions.
Altman Identifies Energy as AI’s Biggest Challenge
Beyond the business agreements, Sam Altman’s appearance at ADIPEC 2026 offered a closer look at the challenges facing OpenAI and the wider AI industry.
In a discussion with Mina Al-Oraibi, editor-in-chief of The National, Altman described energy as the biggest challenge facing the industry. He pointed to the rapid growth in computing demand as AI models become more widely used.
Although the energy required for each individual query has fallen significantly, total consumption continues to rise as usage expands. To illustrate the scale of this growth, Altman compared token consumption over time. He said OpenAI’s largest internal user consumed around 100,000 tokens per month six years ago, a figure that is now close to the average monthly consumption per person worldwide. Today’s heaviest users, he added, can consume around 500 billion tokens per month.
Altman said energy availability and the physical infrastructure needed to operate data centers will remain major factors shaping the cost and scale of AI adoption, even as models become more efficient.
He also argued that AI companies should build stronger relationships with the energy sector. In his view, partnerships with companies such as ADNOC create opportunities not only to secure the resources needed for AI infrastructure, but also to contribute to solving complex challenges within the energy industry.
Altman Discusses AI Progress and a 10,000-Agent Research Project
Altman also addressed the rapid improvement in AI model capabilities over the past year. He said that models had moved from performing below the level of the world’s leading scientists at the beginning of the year to surpassing them in most fields, according to his assessment. He added that AI systems can now handle substantial portions of many office-based jobs.
During the discussion, Altman referred to a recent mathematical achievement involving approximately 10,000 AI agents working simultaneously on one of the Millennium Prize Problems. The agents collaborated for around four days.
His description matches OpenAI’s earlier project involving the Navier-Stokes problem, which the company announced in September 2026. That project demonstrated how large groups of AI agents could work together on a highly complex mathematical research task.
Altman Says Humans Must Remain in Control of AI
Discussing the broader effects of AI on society and employment, Altman stressed that people must retain control over important decisions. He described AI as a tool intended to serve people rather than replace their authority over what should be done.
He also highlighted the importance of demonstrating that AI companies can prevent serious safety incidents and keep increasingly capable models aligned with human values and objectives.
Altman praised the UAE’s approach to AI adoption, describing the country as one of the most advanced in the field. He also estimated that around 80% of young people in the UAE use ChatGPT every week.
Abu Dhabi’s Investment Interest Adds to OpenAI’s Regional Expansion
The four agreements reflect the expanding relationship between OpenAI and UAE organizations, as Abu Dhabi works to develop large-scale computing infrastructure and integrate AI into key economic and public services.
Separately, media reports have pointed to discussions between OpenAI and UAE investment funds, including MGX, over potential participation in a financing round that could value $30 billion. These discussions were reported independently and were not part of the four agreements announced in Abu Dhabi.
Taken together, the ADIPEC 2026 announcements point to a broader shift in how OpenAI is expanding its presence in the UAE. The focus extends beyond making AI models available to individual users, toward embedding them in the operational systems of governments and major businesses. From reviewing imported food shipments and government permits to planning oil field development and analyzing investments, OpenAI and its UAE partners are working to bring AI into the processes that support decisions across critical sectors.
