How do you build a durable culture of responsible AI use rather than just publishing a policy?

Building a durable culture of responsible AI use requires embedding practices into organizational operations rather than relying solely on published policies. Organizations can adopt management-based approaches, such as creating internal risk management systems that include impact assessments, documentation, audits, and continuous monitoring to adapt to evolving AI models and uses [7]. This fosters accountability and proactive governance at the organizational level, ensuring compliance is integrated into daily workflows. Additionally, implementing "policy as code" processes can operationalize guidelines by automating compliance checks during AI adoption, addressing issues like agentic drift and promoting consistent responsible use across teams [2]. Shifting from rhetorical commitments to practical actions, such as data frugality in AI development, helps cultivate a culture that prioritizes efficiency, reduces environmental impact, and sustains long-term responsibility [4].