Jake Moshenko (@jacobmoshenko, CEO of @AuthZed) talks about the challenges of securing AI interactions, as well as authentication best-practices for AI Agents and RAG patterns.  

SHOW: 885

SHOW TRANSCRIPT: The Cloudcast #885 Transcript

SHOW VIDEO: https://youtube.com/@TheCloudcastNET 

CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotw

NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" 

SHOW NOTES:

• Authzed (homepage)
• SpiceDB - Authorization that scales (GitHub)
• Google’s Zanzibar (research paper)
• Annotating Zanzibar (by AuthZed)
• Authzed CEO discusses Auth for AI RAG and Agents (TheCUBE)

Topic 1 - Welcome to the show. Tell us about your background and why you decided to start AuthZed.

Topic 2 - AuthZed is based on technology from Google called Zanzibar. There’s an in-depth research paper, but give us some background on the types of problems it solves and why it was so appealing to you. 

Topic 3 - Authorization is used for every application, but let’s talk about it in the context of AI or AI Agents. This is still a fairly new concept, but what authorization-related challenges do you expect GenAI or AI agents to create for developers and operators?

Topic 4 - Is this space moving so fast that people already realize they’ll likely need something more modern, like Authzed, or are there mistakes they are making today that are going to cause serious red flags soon? 

Topic 5 - What is the typical path for companies or teams to adopt Authzed, and is it something that can start small and grow (or merge with other groups), or does it have to be centralized like at Google? 

Topic 6 - What are some of the more important areas that people building and growing AI systems that need authorization should be keeping an eye on? 

FEEDBACK?

• Email: show at the cloudcast dot net
• Bluesky: @cloudcastpod.bsky.social
• Twitter/X: @cloudcastpod
• Instagram: @cloudcastpod
• TikTok: @cloudcastpod