The missing foundations holding back AI in energy and utilities

Mark Simpson contributed a thought leadership piece to Energy Manager Magazine on the foundations energy and utility organisations need in place before AI can deliver value at scale.

The article covers:

• Why AI initiatives across energy and utilities remain trapped in experimentation, and how the barriers set out in the UK government's vision for an AI-enabled clean energy system, including fragmented data, technical integration, regulatory clarity, skills and culture, hold back adoption

• Why general-purpose models lack critical industry context, since the knowledge that matters most sits in engineering documents, asset databases, legacy maintenance systems and the experience of long-serving employees

• The market shift towards smaller, task-specific AI models, and the risk that generic models miss essential technical, regulatory or operational context in decisions affecting critical infrastructure

• Moving from pilots to enterprise impact by grounding AI in the information employees use every day and prioritising well-defined use cases such as engineering documentation, regulatory compliance and knowledge retrieval for field teams

• Treating explainability and auditability as operational requirements rather than compliance exercises, with clear visibility into data sources and recommendations before AI informs maintenance planning, outage response or network management