This working paper introduces Cultural AI Governance as an interdisciplinary framework for examining how artificial intelligence affects cultural memory, institutional resilience and public meaning. It argues that AI systems operating in cultural and public-interest environments should be assessed not only in terms of technical performance or compliance, but also through representational integrity, accountability and human-centered oversight. A central contribution is the Witness Framework, which moves beyond the narrow notion of the “user” and emphasizes the human subject as an interpretive participant in AI-mediated cultural experience. The paper also connects cultural AI governance with public-sector assurance, impact assessment, auditability and institutional responsibility.
Fotis Zygoulis (Fri,) studied this question.