Lars Meloe, Principal Consultant in DNV’s Digital Trust Venture, Energy Systems, presents a framework for managing risk as organisations develop and use industrial AI. Drawing on work with energy sector partners, he explains why an AI opportunity needs a clear business purpose alongside a plan for safety, resilience, cybersecurity and regulatory requirements.
The framework follows an AI solution from development and delivery into operation. Lars stresses that managing risk cannot end when a system is first approved. AI software and its performance may change over time, so organisations need ways to check that it remains suitable for its intended use throughout its life.
His presentation looks at trust across three connected levels: the AI technology itself, the wider system in which it operates and the context in which people use it. Governance sits across all three. Lars also introduces assurance cases, which test claims about a system against supporting evidence, and a bow tie approach to examining risks and controls.
The central question is how an organisation can show that an AI system deserves trust for a particular industrial task, and continue to support that claim as conditions change. Watch the full presentation to explore the framework and Lars’s explanation of the assurance methods in more detail.
Speaker
- Lars Meloe, Principal Consultant, Digital Trust Venture, Energy Systems, DNV
