Tuesday, August 25|3:30 pm – 4:15 pm
Educational Session|Innovation
Software quality assurance in nuclear settings rests on traceability, repeatability, and verifiable behavior, expectations that AI tools challenge through their non-linear and probabilistic processing. This session explores how practitioners can apply existing SQA methodology to AI-enabled tools and systems while preserving the rigor that safety-critical work demands. Speakers will walk through where current qualification frameworks hold up, where they strain, and how teams are adapting verification, validation, and configuration management to account for systems whose outputs are not fully deterministic.
What You'll Gain
Gain a clearer picture of how AI tools map onto established SQA requirements. Leave with practical approaches for verifying and validating systems that behave probabilistically, for distinguishing AI use cases and the requirement implications, and for maintaining configuration control when models retrain or update.
Who Should Attend
Software QA and V&V engineers, instrumentation and controls engineers, licensing and regulatory affairs professionals, digital and AI deployment leads, and quality assurance managers responsible for qualifying tools and systems in safety-related applications.