905: Why RAG Makes LLMs Less Safe (And How to Fix It), with Bloomberg’s Dr. Sebastian Gehrmann

Episode Link

https://share.snipd.com/episode/5bab1167-646a-4da2-88be-b6ab298a1ff1

Episode publish date

July 15, 2025 11:00 AM (UTC)

Last edit date

Jul 19, 2025 11:54 AM

Last snip date

July 19, 2025 12:54 PM (GMT+1)

Last sync date

July 19, 2025 12:54 PM (GMT+1)

Show

Super Data Science: ML & AI Podcast with Jon Krohn

Show notes link

Snips

Warning

⚠️ Any content within the episode information, snip blocks might be updated or overwritten by Snipd in a future sync. Add your edits or additional notes outside these blocks to keep them safe.

‣

Episode show notes

Your snips

‣

[02:18] RAG Can Reduce Safety Despite Grounding

‣

[09:39] Unique AI Attack Surfaces Exist

‣

[15:18] Guard Data and User Behavior

‣

[18:04] RAG Improves Honesty but Not Harmlessness