Decisions at the Frontier · Episode 7
AI Risk, Regulated Industries & Leading Without Answers
With Ahmad Fattahi, Executive Director, Generative AI
Hosted by Victoria Mensch
A conversation on AI strategy, risk, and leadership.
Watch on YouTubeEpisode summary
How do you make high-stakes decisions about AI when both the technology and the rules around it keep shifting? In this episode of Decisions at the Frontier, Victoria Mensch speaks with Ahmad Fattahi, Executive Director, Generative AI, about AI risk, adoption in regulated industries, and the organizational choices that come with it.
The conversation begins with an analogy: the progression from paper maps to GPS to self-driving cars. It explores how autonomy is earned, then moves into a more nuanced discussion of risk—not simply whether AI is safe or unsafe, but what level of risk is appropriate in a particular setting. Chip manufacturing yields and patient-facing AI provide contrasting contexts for that discussion.
The second half turns to leadership and organizational design. Ahmad and Victoria examine the inverse pyramid and hourglass models, revisit the steam engine and textile factory, and discuss building trust without having all the answers. The closing principle is a reminder that not making a decision is itself a decision.
Key takeaways
Autonomy is earned
The paper-map-to-self-driving-car analogy frames AI as a different kind of technology shift and raises the question of when greater autonomy is justified.
Risk is a spectrum
The discussion moves beyond a black-or-white view of AI risk to the question of what shade of gray a particular use case represents.
Context changes the stakes
Chip manufacturing yields and patient-facing AI illustrate why adoption in regulated, high-stakes environments needs a context-specific discussion.
Organizational shape matters
The episode examines why an inverse-pyramid structure may backfire and explores the case for an hourglass instead.
Trust cannot wait for certainty
Victoria and Ahmad explore what it means to lead and build trust when leaders do not have all the answers.
Inaction is still a choice
The conversation closes with the principle that not making a decision is a decision in itself.
Transcript
A verified written transcript is not yet available here. The summary and takeaways above are based on the episode description, not a verbatim transcript.
View episode on YouTubeContinue the conversation
Explore more executive conversations, or connect these questions to your organization’s leadership learning journey.