Скачать или смотреть Hallucinated Package Exploits—Steve Wilson Live with Tim O'Reilly

Watch the entire conversation: https://learning.oreilly.com/videos/b...


As the founder of the OWASP Gen AI Security Project and the author of The Developer’s Playbook for Large Language Model Security, Steve Wilson is well acquainted with the many things that should be keeping AI-assisted developers up at night. Topping the list is hallucinations. The fact that LLMs often confidently just “make stuff up” is one of the most severe vulnerabilities in these models, says Steve. Take hallucinated package exploits: “The bad guys have done all the research to find the commonly hallucinated packages and they've made nefarious versions of them and put them up on GitHub.” So be careful what you import.

Follow O'Reilly on:
LinkedIn:   / oreilly  
Facebook:   / oreilly  
Instagram:   / oreillymedia  
BlueSky: https://bsky.app/profile/oreilly.bsky...