Dinergy's AI trust check — a single question posed at the end of every Weekly Pulse — has become one of the most discussed governance innovations in the energy sector. The protocol requires every participant to answer honestly whether any artificial intelligence output produced an unexpected or unexplained result during the preceding week.

The concept is deceptively simple. At the conclusion of every Weekly Pulse briefing, after the four dimensions of review have been completed and all action items assigned, the Founder Architect poses a single question: "Did any artificial intelligence output surprise you this week?"

If the answer is yes — from any participant — the event is logged, categorized, and assigned for investigation. The investigation must determine whether the surprise was a positive innovation (the AI found a pattern humans missed), a neutral anomaly (the AI produced an unusual but harmless output), or a concerning deviation (the AI behaved in a way that suggests a flaw in its model or training data).

"This is not about being afraid of AI," Merlin has explained. "This is about being honest about AI. If we cannot explain why an AI system produced a particular output, we have a governance gap. And governance gaps in energy systems are not theoretical risks — they are operational risks with real consequences."

The AI trust check has produced several valuable insights since its implementation. In one case, an AI agent responsible for trading recommendations produced a series of unusually conservative position suggestions during a period of market volatility. Investigation revealed that the agent's training data included a period of extreme market stress that was causing it to over-weight downside scenarios. The model was recalibrated, and the incident was recorded in the decision log as an architectural decision record.

The protocol has attracted attention from other organizations in the energy sector, several of which have adopted similar practices. Dinergy has published the framework as an open governance standard, consistent with its philosophy that AI governance should be a shared responsibility across the industry.