ADNOC Says It Runs More Than 200 AI Applications and 70 Robotics Projects

Al Hashmi said employees across the group have also built thousands of AI agents themselves. The aim, she said, is to make the technology part of everyday work instead of leaving it to specialist technology teams. According to Al Hashmi, more than 40,000 ADNOC employees have received AI training over the past two years.

"It is not about applying AI everywhere. It is about applying AI where we can see the value," she said. She pointed to gains in efficiency and production, and to better prediction of equipment failures.

Where AI Is Being Used

At ADNOC Offshore, an intelligent operating centre uses AI to monitor operations and warn frontline teams about problems before they hit production, according to the report. ADNOC has previously described robotics work across drilling, production, infrastructure, logistics and shipping. This includes robots and drones that inspect facilities, monitor pipelines, detect emissions and work in hazardous areas.

In May the company deployed a heavy-duty inspection robot at its Taweelah Gas Compression Plant. The robot carries thermal cameras and sensors and can inspect industrial areas on its own, looking for gas leaks and hotspots. ADNOC is also developing an operator robot that can handle equipment, turn valves and operate gauges, work that would normally require staff to enter dangerous areas.

Al Hashmi said AI will play a growing part in operational decisions but people will keep final responsibility. "Technology will not transform our industry. People do," she said.

Oilfield Service Chiefs Flag Security Risks

Other executives on the panel focused on the risks. SLB chief executive Olivier Le Peuch said a key concern was whether the data fed into AI systems could be trusted. He warned that compromised data could cause problems as companies build specialised models for critical industrial processes. He also said AI could draw value from decades of exploration and production data that has never been fully used, but that scientific and engineering standards could not be compromised.

Halliburton chief executive Jeff Miller said cybersecurity now reaches the systems that run physical equipment, not just corporate IT networks. "What you might have thought is an IT issue is actually an OT issue," he said, referring to operational technology, the systems that monitor and control industrial plant.

The OpenAI and G42 Deal

The figures came a day after ADNOC signed a multi-year AI agreement with OpenAI and G42 when ADIPEC opened on Monday. Gulf News reported that Sultan Ahmed Al Jaber, UAE Minister of Industry and Advanced Technology and ADNOC Managing Director and Group CEO, said the deal would transform the company's field development, asset management and commercial operations. He said it would also make ADNOC the world's first energy company to deploy OpenAI frontier.

According to Khaleej Times, the agreement is meant to link AI-driven decisions across production, processing, logistics, sales and trading. That would let ADNOC weigh operational and commercial decisions together instead of handling each part of the business separately.

OpenAI chief executive Sam Altman attended the opening. Gulf News reported that he named energy as the biggest challenge for AI and said that "AI needs to serve people, not the other way around."

The ADNOC figures are one of the more concrete public measures of how far a Gulf national oil company has gone with AI. They show the technology moving from analysis into decisions about physical assets. The warnings from SLB and Halliburton show what that move costs: the more AI is connected to industrial control systems, the more the reliability of data and operational networks matters.