This story is from November 18, 2026. For the latest, see today's news.
November 18, 2026 · AI Tech
In a new NVIDIA AI Podcast episode, Hugging Face co-founder Thomas Wolf and ServiceNow's Amanda Grady, SVP of AI Platform Security, explore how to secure when traditional safeguards—like asking models to behave—are no longer sufficient.
They discuss why firm technical boundaries work better than persuadable models, lessons from an attack on Hugging Face using a swarm of AI agents, and the concept of "AI Passport" identity systems. The conversation also covers why open-source approaches can actually improve security rather than weaken it, and how security enables companies to deploy agents in real-world applications.
Topics include physical AI, accessibility, and who should be able to build AI systems. The full episode is available on major podcast platforms.
This is an AI-written summary. The original reporting belongs to @NVIDIA. How we write these · Spot a mistake?
Get the AI brief every morning.
The five AI stories that matter each morning: startups, money, jobs, tech and what's coming next. Free, and one click to unsubscribe. See our Privacy Policy.