Salesforce’s Chief Ethical and Humane Use Officer, Paula Goldman, joins Ross Dawson to discuss how organizations can design and deploy AI responsibly while maximizing its benefits. She emphasizes that ethical AI is not just about avoiding harm, but about fostering trust through transparency, safeguards, and collaborative design. Goldman outlines practical strategies like adversarial testing, inclusive stakeholder input, and clear escalation processes between humans and AI to ensure systems perform reliably and support meaningful human oversight. She also introduces the concept of “human at the helm,” reframing traditional “human in the loop” models to better suit autonomous, agentic AI while preserving human control over critical decisions.

The conversation also covers AI governance in real-world settings, advocating for iterative, minimum viable governance structures that start with clear goals, basic performance metrics, and heightened diligence for sensitive use cases. Goldman stresses the importance of board-level oversight, diverse advisory councils, and open channels for employee feedback to ensure responsible deployment. Ultimately, she envisions a future where AI augments human creativity, deepens connections, and frees people for higher-value work—underscoring that the technology’s positive potential depends on intentional, values-led implementation.