Ignite: Security of Open-Source AI: is there any difference?

on YouTube ↗05m

" summary

Claims that application and infrastructure account for 80% of AI security risks and that classic infosec controls mitigate most exposure. Shows a Hugging Face model labeled “totally harmless” that includes a pickle file made it malicious, and cites a popular GitHub repo with an OpenSSF score 3.7. Notes OAS LLM top‑10 docs omit open source explicitly, recommends implement non‑AI IT and privacy controls first, define internal AI policy and run an open‑source program to reduce real risk.

AI-written summary · about

" related talks

• 21mNavigating the Intersection: AI’s Role in Shaping the Secure Open Source Software Ecos...Harry Toor · CloudNativeSecurityCon North America 2024
• 17mLightning Talk: Navigating the Intersection: AI’s Role in Shaping the Secure Open Sour...Harry Toor · Secure AI Summit 2024 (Powered by Cloud Native)
• 14mDesigning the open and safe AI futureAnne Bertucio · Gemma Developer Day 2024
• 16mKeynote: Open Source for AI and AI for Open SourcePriya Nagpurkar · Open Source Summit Korea 2026
• 36mThe Dark Side of AI: The Hidden Risks in Open-Source AI ModelsTal Folkman · Open Source Summit Japan + AI_dev 2024
• 42mThe Open Source AI Dilemma: Crafting a Clear Definition for Open Source AIOfer Hermoni, Mer Joyce · Open AI + Data Forum @ Open Source Summit North America 2024
• 37mThe Open Source AI Dilemma: Crafting a Clear Definition for Open Source AIOfer Hermoni · AI_dev: Open Source GenAI & ML Summit Europe 2024
• 29mThe Open Source Imperative in AI: Transparency, Governance, and Global CollaborationC. Xie, J. McCrosky · Open Source Summit Korea 2025