# Fiduciary Duties and AI Legal Frameworks, Technical Implementation, and Governance This is a cleaned Markdown copy of the attendee information file for the workshop. The embedded map image from the original file is omitted here to keep the public site simple. ## About This Workshop This two-day workshop explores the emerging intersection of fiduciary duties and artificial intelligence systems. As AI agents increasingly act on behalf of individuals and organizations in high-stakes domains - managing health information, making financial decisions, curating educational content, and mediating consumer relationships - questions of loyalty, care, and accountability become paramount. This conference aims to bootstrap an interdisciplinary community of practice that will shape the development, regulation, and standardization of fiduciary AI systems. We seek to bring together legal scholars, computer scientists, entrepreneurs, civil society organizations, and government representatives to examine how traditional fiduciary principles can be adapted, implemented, and enforced in the age of AI agents. The workshop takes up a set of open questions: whether AI systems today are capable of performing a fiduciary role, and how we will know when they are ready; how they can be designed to comply with existing fiduciary duties across law, healthcare, finance, and the guardianship of children; what new responsibilities should apply to AI providers, and which actors in complex AI supply chains bear them; and what business models will sustain a thriving ecosystem of fiduciary AI services. Throughout, it asks which institutions will enforce these duties and how. We will meet using Chatham House rules to enable attendees to speak with more candor. Dates: June 4-5, 2026 Venue: Vanderbilt Hall, New York University School of Law, 40 Washington Square S, New York, NY 10012 Co-organized by: NYU School of Law and GliaNet Alliance Sponsored by: Future of Life Foundation (FLF) ## Day 1: Thursday, June 4 Greenberg Lounge, NYU Law School | Time | Session | | --- | --- | | 8:30 - 9:00 AM | Breakfast and Registration | | 9:00 - 9:20 AM | Opening Remarks - Sebastian Benthall, NYU Law | | 9:20 - 9:35 AM | Lightning Round Introductions | | 9:35 - 10:00 AM | The Challenge of Being a Fiduciary AI - Richard Whitt and Nell Clasby, GliaNet Alliance | | 10:00 - 10:25 AM | Encoding Agency Law Principles into AI Agent Behavior Evaluation - Dan Leininger and Daniella Raposo, Consumer Reports | | 10:25 - 10:40 AM | Break | | 10:40 - 10:55 AM | The Fiduciary Gap in Agentic AI: What Enterprise Governance Practitioners Are Seeing and Building - Ken Priore, Docusign | | 10:55 - 11:10 AM | Using Client Facing AI to Provide Investment Advice - Kate Ring, Stash | | 11:10 - 11:30 AM | Reflection and Open Discussion on Perspectives from Practice - facilitated by Richard Whitt, GliaNet Alliance | | 11:30 - 11:45 AM | AISPA: Artificial Intelligence System Prompt Assurance - Jiaxin Pei, Stanford University | | 11:45 AM - 12:00 PM | Designing AI Systems for Fiduciary Contexts: Runtime Assurance, Oversight, and Delegation - Mateo Petel, Stanford University | | 12:00 - 1:00 PM | Lunch | | 1:00 - 1:15 PM | Loyal Agent Evals: A Benchmark Framework for Measuring AI Compliance with Contractual Fiduciary Duties - Dazza Greenwood, Civics.com / Stanford Digital Economy Lab | | 1:15 - 1:30 PM | How to Choose a Threshold for an Evaluation Metric for Large Language Models - Dhagesh Mehta, BlackRock | | 1:30 - 1:50 PM | Reflections and Open Discussion on The Future: Technical Research I and II - facilitated by Sebastian Benthall, NYU Law | | 1:50 - 2:40 PM | Aligning Incentives: Business Models for the Fiduciary - John Dewees, Blue Orb, Inc. and Richard Whitt, GliaNet Alliance | | 2:40 - 3:00 PM | AI Loyalty - Neil Richards, Washington University in St. Louis | | 3:00 - 3:20 PM | Reflections and Open Discussion on Governance Policy Implications for Supporting a Fiduciary - facilitated by Richard Whitt, GliaNet Alliance | | 3:20 - 3:35 PM | Break | | 3:35 - 4:35 PM | Law, Public Policy, and Institutional Interventions for Fiduciary AI - facilitated by Sebastian Benthall, NYU Law | | 4:35 - 4:55 PM | Day 1 Closing Remarks - facilitated by Sebastian Benthall, NYU Law | | 5:30 - 8:30 PM | Networking Reception hosted by GliaNet Alliance at 310 Bowery Bar NYC, 310 Bowery, New York, NY 10012 | ## Day 2: Friday, June 5 Greenberg Lounge, NYU Law School Friday includes parallel sessions across two tracks. Track A is in Vanderbilt Hall Room 202. Track B is in Room 208. Plenary sessions will be held in Greenberg Lounge. ### Morning Plenary | Time | Session | | --- | --- | | 8:45 - 9:10 AM | Breakfast | | 9:10 - 9:20 AM | Orienting Remarks | ### Paper Sessions: Parallel Tracks Track A will be in Vanderbilt 202. Track B will be in Vanderbilt 208. | Time | Track A - Automation, Complexity, and Liability | Track B - Humans, Judgment, and Delegation | | --- | --- | --- | | 9:20 - 10:30 AM | How can fiduciary roles be automated? How can liability be determined in AI supply chains? Presenters: Ben Sundhold, Tulane; Marco Germano, University of Toronto; Jiaying Jiang, University of Florida. Discussion moderated by Mateo Petel. | When and how can human fiduciaries delegate their decisions to AI? Presenters: Kavell Joseph, World Bank; Genevieve Helleringer, Oxford; Ramona Afroze, Widener. Discussion moderated by Brett Frischmann. | | 10:30 - 10:50 AM | Break | Break | | 10:50 AM - 12:00 PM | AI Fiduciaries and their Discontents: When should AI systems be fiduciaries? Are fiduciary duties too weak, or too strong, for AI? Presenters: William Marks, Harvard BKC; Christina Lee, UC San Diego; Blaine Dillingham, FAI. Discussion moderated by Neil Richards. | Roles, Duties, and Prerogatives: What do human relationships - family and friendship - have to teach us about fiduciary AI? Presenters: Nina Russell, UC Berkeley; Brett Frischmann, Villanova; Andrew Gold, UC Irvine. Discussion moderated by Sebastian Benthall. | ### Midday Plenary | Time | Session | | --- | --- | | 12:00 - 1:00 PM | Lunch. Table themes: ex post enforcement; technical implementations for authority and apparent authority; best execution rules. | | 1:00 - 1:10 PM | Mapping the Landscape: Introducing the activity. We will create maps of actors, duties, technologies, and institutions to guide strategic action. Breakout groups will have different foci and different agentic AI tools. What are our confident next actions? What are our open questions? | ### Breakout Groups | Time | Group A - AI in Regulated Sectors | Group B - Creating New AI Fiduciaries | Group C - Agentic AI Collaboration | | --- | --- | --- | --- | | 1:10 - 2:15 PM | Focus on implementing AI in regulated sectors and roles. | Focus on fiduciary AI in domains that currently do not have legally established duties. | Using agentic AI to map the path forward for fiduciary AI. [Interlateral](https://interlateral.com/). AI agent required. Prepare at lunch table. Meets in Greenberg Lounge. | ### Afternoon Plenary | Time | Session | | --- | --- | | 2:15 - 3:15 PM | Mapping the Landscape. Each group presents their map to the other. General discussion of open problems. | | 3:00 - 3:15 PM | Break | | 3:15 - 4:15 PM | Movement Building facilitated by Ben Moskowitz. We will discuss our next actionable steps for making fiduciary AI a reality. | ## Logistics Venue: Vanderbilt Hall, NYU School of Law, 40 Washington Square S, New York, NY 10012. The workshop will be at Greenberg Lounge, just down the hall from the building entrance on the first floor. The security guard will know of the event at Greenberg Lounge. Participants should have a photo ID in case security asks for one. ## Networking Reception Thursday, June 4, 5:30-8:30 PM, hosted by GliaNet Alliance at Bowery Bar NYC, 310 Bowery, New York, NY 10012. ## Paper Drafts and Presentations Discussion drafts of papers being presented at the workshop are available in this [shared Drive folder](https://drive.google.com/drive/folders/15Te_madD3vLDBkxqE2jgw80Fyg-CEeQw?usp=drive_link). If you would like to add or change your draft, please write to Sebastian Benthall at [spb413@nyu.edu](mailto:spb413@nyu.edu). You are encouraged to read the drafts before arrival. If you will be presenting slides and wish to share your work with the workshop community, please add those to [this shared Drive folder](https://drive.google.com/drive/u/1/folders/1dAdu8hD-7OrB9YOyjZGWPBiAHTXLqmes). ## Contact For questions about the program: Sebastian Benthall, NYU, [spb413@nyu.edu](mailto:spb413@nyu.edu). For logistics and the GliaNet-hosted reception: Estefanie J. Govea, [estefanie@glia.net](mailto:estefanie@glia.net).