Future Oil & Gas 2026: the sector no longer asks "should we use AI?" but "how do we scale it safely?"

Meeting in Aberdeen on 24-25 June 2026, 264 delegates from 174 organisations confirmed a quiet but decisive shift: the oil and gas industry has left the experimentation phase behind. Around the table, major operators — Shell, bp, ExxonMobil, Aker BP, MOL — and a specialist vendor ecosystem — Falkor (formerly Kongsberg Digital), Enablon, KBC, VEERUM, Prometheus Group — were no longer debating the theoretical merits of AI, but the concrete conditions for deploying it at scale. Adam Soroka, CEO of Cavendish Group, summed up the year: "the question is no longer whether we are going to apply AI, but how we are actually applying it."

This shift changes the nature of the problem. As long as AI stays at pilot stage, the challenge is technical: one model, one dataset, one demonstration. The moment you generalise across dozens of sites, assets and workflows, the challenge becomes organisational: data governance, decision traceability, upskilling field teams, and above all control structures able to evolve as fast as the technology. Speakers were blunt: without a robust governance framework, scaling does not accelerate value — it multiplies risk.

For a sector where every critical software layer touches facility safety, the environment and regulatory compliance, this requirement is not a nice-to-have: it is the price of entry. Rising energy demand and decarbonisation pressure force speed, but operational accountability forbids doing it without a safety net. The real differentiator of 2026 is therefore not the highest-performing model, but the ability to industrialise a deployment — data, processes, people — in an auditable and reversible way.

Cardan-AI Analysis: what oil & gas is living through today, aerospace, defence and energy are living through in parallel. The bottleneck has moved from technical feasibility to organisational execution, and that is where ROI is either captured or diluted. Our conviction: before adding yet another use case, map the critical data, set governance rules proportionate to risk, and equip teams to hold the pace without losing control. It is less spectacular than a successful pilot — but it is what separates a demo from an asset that produces value year after year.