Artificial intelligence solutions are becoming increasingly integrated across the oil and gas value chain, according to research by BMI.
However, adoption rates remain uneven. International majors and selected national oil companies (NOCs) are leading deployment, while smaller independent producers and smaller NOCs report lower implementation rates.
Current use cases span upstream and downstream operations. Upstream deployment focuses on subsurface engineering and drilling optimisation workflows. Downstream applications primarily target predictive maintenance and emissions monitoring.
Several producers have tied AI implementation to corporate financial targets. ExxonMobil, in partnership with Halliburton, deployed a fully automated drilling process in offshore Guyana that reduced tripping operations time by 33 per cent and completed its reservoir section 15 per cent ahead of schedule. The company subsequently raised its cumulative structural cost savings target to USD 20 billion by 2030 (versus 2019 levels), alongside plans to double Permian Basin production to 2.5 million barrels of oil equivalent per day.
Shell aims to deliver cumulative structural cost reductions of USD 5 billion to USD 8 billion between 2022 and 2028. These targets depend on procurement and supply chain optimisation, corporate simplification, and technology and AI deployment. Among independents, Devon Energy achieved over USD 250 million in efficiency gains in 2025 using drilling-focused AI agents and models. Saudi Aramco reported USD 2.6 billion in AI technology realised value for 2025, a 33 per cent increase in technology-related savings over 2024.
AI adoption is driven primarily by majors — including Shell, ExxonMobil, Chevron, BP, and TotalEnergies — partnering with oilfield service providers (SLB, Halliburton) and technology developers (Microsoft, NVIDIA). Capital-abundant NOCs like Saudi Aramco, ADNOC, and Petrobras report adoption levels comparable to private majors.
Conversely, smaller independent producers are constrained by capital budgets. This divergence is reflected in the Q2 2024 Dallas Fed Energy Survey of exploration and production operators in Texas, New Mexico, and Louisiana, which measured companies with no near-term plans to utilise AI:
-
Small E&P companies: 57 per cent neither use nor plan to use AI.
-
Large E&P companies: 31 per cent neither use nor plan to use AI.
While deployment is accelerating, BMI outlines three systemic variables that may delay widespread market uptake:
-
Regulatory frameworks: Tightening emissions-reporting rules encourage technical innovation. However, stricter safety, leak-prevention, and unplanned outage compliance protocols may restrict the integration of unproven autonomous systems.
-
Cybersecurity risks: Broader digital networks increase exposure to cybersecurity threats. Consequently, corporate expenditure on IT security is projected to rise to prevent operational disruptions.
-
Technical skill deficits: The sector faces an internal shortage of advanced AI technical expertise.