Energy: the sector most convinced by AI is also the most lost in measuring it

EY's US AI Pulse Survey, Wave 4, published April 13, 2026 from data collected in December 2025, puts the energy sector at the top of a rarely-highlighted metric: organizational interest in responsible AI has jumped 25 points since July 2024, reaching 72% of senior leaders — the highest of any industry. Among leaders whose organization invested in AI and saw productivity gains, 78% say those gains are a catalyst for strategic transformation, versus 55% cross-industry.

The same survey reveals a contradictory signal: 56% of energy leaders say they struggle to directly link productivity improvements to their AI investments, versus 35% cross-industry — a 21-point gap. And 72% say they need more training to accurately communicate those productivity gains, versus a 50% cross-industry average.

The sector with the strongest conviction that AI is transforming its strategic model is also the one struggling most to prove it with numbers. This is not an incidental contradiction: it is the signature of a measurement problem, not a value problem — conviction is running ahead of evidence, and the gap between the two widens exactly where enthusiasm is strongest.

For Cardan-AI, this mismatch has a direct implication for energy and O&G leadership: before allocating capital across AI projects based on their reported impact, organizations first need to invest in measurement infrastructure — causal attribution, pre-deployment baseline metrics — or scaling decisions will be made on conviction rather than evidence.