Episode 69: Anthony Katsur, CEO at IAB Tech Lab: Navigating AI, Privacy, and Adtech’s Agentic Future

Tony Katsur, CEO of IAB Tech Lab, joins Alan Chapell to unpack the rise of agentic AI and its impact on privacy and data governance. They explore why current approaches fall short and how the industry risks repeating past mistakes. The discussion breaks down vector embeddings and how they enable data matching without exposing raw user information. They also examine ongoing challenges around compliance, consent, and data deletion in AI-driven systems. Finally, the episode covers AI content marketplaces and how frameworks like COMP aim to bring structure, transparency, and fair compensation to publishers. The Chapell Regulatory Insider may be found here: https://chapellreport.substack.com/ Takeaways Agentic AI brings powerful automation but risks repeating past privacy failures without strong governance frameworks in place. Existing privacy standards like TCF and GPP can be embedded into agentic systems but need further evolution and enforcement. Vector embeddings enable privacy-conscious data matching by comparing similarity rather than sharing raw data. Data deletion and compliance remain unresolved challenges when user data is embedded into AI models or vectors. Audit, attestation, and accountability mechanisms are critical to prevent misuse and misrepresentation in agentic ecosystems. AI content marketplaces require structured licensing frameworks like COMP to support fair compensation and transparency. Tokenization of content could improve tracking, attribution, and source-of-truth verification for publishers and brands. The industry is still early in agentic development and must slow down to build privacy-first foundations. Chapters 00:00 Intro and discussion on agentic AI hype vs reality 01:12 Why privacy is missing from the agentic AI conversation 03:34 Challenges with DSARs and scaling privacy compliance 05:08 Existing privacy frameworks and how they apply to agentic systems 07:44 The role of privacy taxonomy and data classification 12:38 Explaining vector embeddings and privacy-safe data matching 18:30 Compliance challenges with embeddings and data transparency 24:26 Agent registry and identity verification in agentic systems 30:54 AI content marketplaces and the COMP framework 35:24 COMP vs RSL and licensing models 38:15 Content tracking, tokenization, and transparency challenges 42:48 The future of AI content marketplaces 44:38 Why industry participation in Tech Lab is critical Learn more about your ad choices. Visit megaphone.fm/adchoices