Booby Trapping LLMs - AI Frenzy in Cybercrime

Hosts: Steve Gibson and Leo Laporte Security Now episode 1100 More Info: https://twit.tv/shows/security-now/episodes/1100 Sponsors: • http://threatlocker.com/twit • http://hoxhunt.com/securitynow • http://originhq.com/securitynow • http://trustedtech.team/twit Think your open-source AI model is safe? This episode uncovers how LLMs can be secretly booby-trapped with undetectable backdoors, raising the stakes for anyone deploying AI in serious environments. • Anthropic cuts live Internet access during testing. • Will LLM technology turn out to be an uncontrollable dead end? • What alternative AI technologies are underway? • Anthropic offers free AI vulnerability scanning for open source. • South Korean banks hit by Claude-driven cyberattacks. • Japan declares a state of cyber-emergency due to attacks. • For one listener, agentic coding has taken the joy out of coding. • A listener is having way too much fun with AI and Halloween. • Some listeners believe I'm downplaying the dangers of AI. • We carefully examine the consequences of maliciously trained AI Show Notes - https://www.grc.com/sn/SN-1100-Notes.pdf Products we recommend: https://www.amazon.com/shop/twitnetcastnetwork TWiT may earn commissions on certain products. Follow us: Website: https://twit.tv/ Bluesky: https://bsky.app/profile/twit.tv Mastodon: https://mastodon.social/@TWiT Facebook: https://www.facebook.com/TWiTNetwork Instagram: https://www.instagram.com/twit.tv/ X (Twitter): https://x.com/TWiT TikTok: https://www.tiktok.com/@twittok Reddit: https://www.reddit.com/r/thisweekintech/ LinkedIn: https://www.linkedin.com/company/twit-llc About us: TWiT.tv is a technology podcasting network located in the San Francisco Bay Area with the #1 ranked technology podcast This Week in Tech hosted by Leo Laporte. Every week we produce dozens of hours of content on a variety of programs including Tech News Weekly, MacBreak Weekly, Windows Weekly, Security Now, Intelligent Machines, and more.