AI and IoT security - The road ahead. Expert Panel
summary
Claim that AI-infused zero-days shift from one or two per year to multiple per week and get brokered by nation-states is repeated, with a concrete warning that prompt injection is nearly impossible to detect and agentic AIs get delegated real authority. Concrete mitigations named include memory-safe languages, compartmentalization, PQC and QKD for a 2030–2035 quantum threat, AI-driven SOCs and agent inventories, and content provenance standards like C2PA to label model‑generated media.
AI-written summary · about
chapters (7)
- 0:00Welcome
- 1:05Panelist backgrounds
- 4:59Risk landscape changes
- 9:56AI-driven defense
- 17:24Human role in cybersecurity
- 26:36Deep fakes and code review
- 31:19AI and IoT security future
related talks
20mIoT Security: Past, Present and Future
49mThe Intersection of AI and Cybersecurity
42mAI for security - friend or foe?
22mFireside Chat - AI Security Vulnerabilities
37mFundamentals of AI safety and security
13mAI - Out of Control
20mSecurity, and AI... so happy together?
31mSecuring the Future: Navigating AI Vulnerabilities and Evolving Security Practices