Leading AI Gov: Governance for Multimodal AI Text, Vision, Voice, and Beyond

Most governance programs were built around structured data and human decisions — but multimodal AI is ingesting voice, video, and unstructured content and drawing inferences your policies were never designed to cover. This session gets specific about where existing governance frameworks hold up and where they break down. Kelle O'Neal and Lisa Wintrick of First San Francisco Partners walked through their framework for governing multimodal AI, organized around two planes: the ingestion plane (what AI is allowed to take in) and the inference plane (what it is allowed to conclude). They used meeting transcription bots as a recurring case study to show how tools can collect biometric data, infer protected characteristics, and create datasets your procurement team never accounted for. The session revisited the pod accountability model introduced in an earlier installment — AI engineer, decision steward, data steward, business SME, and AI communications lead — and applied it specifically to multimodal deployments. It also covered how to bring privacy, compliance, and ethics together at the moment of decision, and closed with concrete talking points for board-level AI governance conversations. Most relevant for data governance leads, AI governance practitioners, chief data officers, and risk and compliance professionals navigating AI deployments in production. Download slides: https://content.dataversity.net/rs/65... More Leading AI Governance on-demand webinars: https://www.dataversity.net/resources... Upcoming Leading AI Governance webinars: https://www.dataversity.net/events/?_... View all upcoming online events: https://www.dataversity.net/events/?_...