Securing AI: From Models to Agents

Est. audience

900

Cadence

Daily

Latest

2026-06-12

By

Honghui Xu's IntelliTrust Lab

Based on

12 recent episodes

• Covers: Attacks, defenses, and architectural security for AI models, multimodal LLMs, and autonomous agents.
• For: AI researchers, security engineers, and developers building secure AI systems.
• Best for: Engineers and technical leaders deploying and securing production AI systems.
• Skip if: Non-technical professionals looking for general business or non-security AI news.

Start with

• Industry Reality: What Big Tech Is Actually Doing: Real-world AI security practices, challenges, and lessons from big tech industry deployments.
• Build a Secure Physical AI System under Real-world Constraints?: Balancing security, usability, cost, and performance when deploying physical AI systems.
• Defense Evolution: From Prompt Tricks to Architectural Security: Why prompt tricks fail and how to implement stronger architectural security in AI.

Visit

Summarised by AI from 12 recent episodes. Audience estimate and Apple rating from Podscan, checked 2026-11-03; treat the estimate as a guide and ask the show for its own numbers. Prices are not included. Report a mistake.