ChatGPT, Claude, North: 3 LLM Developers and Key Perspectives on AI and Privacy

This panel will explore how large language model developers are responding to these new obligations, and whether they align with their own priorities for responsible governance, account for different distribution methods and contexts, and allocate responsibility appropriately. LLMs have captivated the attention of policymakers, but there has been less focus on how safeguards have improved, including reduction in hallucinations and enhanced privacy protections. At the same time, individuals are sharing more data with LLMs, requiring companies to examine how to continue to balance privacy and user safety. Against this backdrop, policymakers have proposed and, in some instances, passed legislation such as the EU AI Act’s training transparency template, which requires detailed disclosures, and systemic risk model analyses, and U.S. state frontier model legislation. Panelists will also discuss LLM hot topics, such as how privacy safeguards work in practice, including operationalizing deletion requests, and downstream information sharing to customers.

What you will learn: 

• The current state-of-play with respect to privacy and security safeguards that LLM developers implement and how they are operationalized. 
• The impact of recent legislative requirements, such as the EU AI Act and New York's RAISE Act, if signed into law, and whether the governance approaches map to the practical operation of LLMs, distribution models, and different contexts such as B2B.
• How to address key issues business customers face.