Andrew Ng

AI Dev 25 X NYC | Panel: Building Trustworthy AI Through Governance, Literacy, and Community

The AI Dev 25 x NYC panel centered on how the industry can rebuild public confidence in artificial intelligence by focusing on three pillars: robust governance, widespread AI literacy, and an engaged community. Miriam, the author of a new book on trustworthy AI, opened the discussion by highlighting the paradox that while 80‑90% of firms now deploy AI, fewer than 11% have formal governance structures in place, creating a fertile ground for mistrust. Speakers emphasized two critical gaps. First, governance must move beyond high‑level principles to concrete, auditable processes—transparent model cards, continuous retesting, and clear accountability at every stage of deployment. Second, AI literacy is essential not just for engineers but for the broader public; people need to recognize how often they already interact with AI and understand its limits. The panel cited a McKinsey 2025 study showing that C‑suite, especially CEO, commitment to AI governance is the strongest predictor of successful generative‑AI adoption, and they advocated sandbox environments where teams can experiment safely before scaling. Miriam’s book outlines nine categories of AI risk, from privacy to hallucinations, and the conversation referenced real‑world examples: Anthropic’s rigorous red‑team testing versus other firms’ more lax documentation, California’s pioneering frontier‑model law, and the stalled federal moratorium on state AI statutes. Andrew warned that over‑regulation or sensationalist media coverage can backfire, stifling innovation while amplifying fear, and stressed that responsible governance should be a catalyst, not a barrier, to business growth. The takeaway for executives is clear: integrating trustworthy‑AI practices is no longer optional. Companies that embed governance into product pipelines, invest in internal AI education, and navigate the evolving regulatory patchwork will protect brand reputation, mitigate legal exposure, and unlock the commercial upside of generative AI. Conversely, firms that ignore these imperatives risk losing consumer trust and falling behind competitors who can demonstrate responsible AI stewardship.