6 Oct 2026
Who's Liable When AI Goes Wrong? What Leaders Need to Know About AI Safety | Dr. Tim Marple
In this episode of the AI Lab Podcast, hosts Ryan Kurt and Mary Grace Glascott sit down with Dr. Tim Marple, Co-Director of Maiden Labs and former frontier AI risk lead at OpenAI and Google. They discuss why AI safety has suddenly become a top concern in the news and in executive forums, and what business leaders should actually do about it.
Tim offers an alternate reading of the recent Hugging Face incident, where AI agents being tested by OpenAI escaped their test environment. He argues it was as much a failure of organizational competence as of model capability. The conversation moves from the headlines to the boardroom: who is liable when AI goes wrong, why most enterprise contracts leave that risk with the customer, and why the insurance industry may be the one to force the issue. Tim also shares a simple framework leaders can use to evaluate AI risk in their own organizations. He makes the case that safety and sound economics are closely tied.
We also dive into:
What the Hugging Face incident reveals about how frontier AI labs manage risk
Why AI liability is still unpriced, and what your MSA may not cover
The "Three C's" of AI risk: competence, character, and context
Why the market needs better ways to judge AI claims, and how to be a discerning AI customer
How being model-agnostic connects cost savings to safety
Reasons for hope, from a growing market for AI safety to a generation of AI-native leaders
Guest Bio:
Dr. Tim Marple is Co-Director of Maiden Labs, a nonprofit research lab that gives industry leaders, labor organizations, and the public data-driven insight into the real-world risks of generative AI. Before joining Maiden Labs, Tim created the first frontier AI risk assessment systems at both Google and OpenAI. There he built and led programs to identify, prioritize, and mitigate risks from AI models and products. His research focuses on the global governance and political economy of emerging technologies and has been published in eight academic and policy journals. Tim serves on the editorial board of Business and Politics, is an associated researcher with the Berkeley APEC Study Center, and teaches an MBA course on technology and risk. He holds a Ph.D. in Political Science from the University of California, Berkeley, and a B.A. and M.A. from the University of Massachusetts Amherst.
About Ryan Kurt:
Ryan Kurt is the Founder of AI Lab, a community and advisory platform built to help executives navigate generative AI. After experiencing layoffs firsthand, he built AI Lab to help leaders and teams reskill for the AI era—equipping organizations to grow through AI, not just cut costs.
About Mary Grace Glascott:
Mary Grace Glascott is the Head of Community at the AI Lab, leading member engagement and content efforts. Prior, Mary Grace was at Google, where she led product strategy and market intelligence as part of Google Cloud’s Applied AI Product Management team, specifically Gemini Enterprise. Mary Grace has also led WW partner marketing at Tableau through its acquisition by Salesforce, was the founding product marketer at the AI start-up Narrative Science, and launched Watson Analytics along with other predictive analytics and BI products while at IBM.
Resources:
→ Learn more about AI Lab: theailab.ai (http://theailab.ai)
→ Connect with Ryan: https://www.linkedin.com/in/ryankurt/ (https://www.linkedin.com/in/ryankurt/)
→ Connect with Mary Grace: https://www.linkedin.com/in/mgbateman13/ (https://www.linkedin.com/in/mgbateman13/)
→ Send listener questions: podcastquestions@theailab.ai (mailto:podcastquestions@theailab.ai)
