IoT Security: Past, Present and Future
summary
Argues that generative AI brings measurable benefits but requires concrete defenses: models are attackable at every lifecycle stage, NIST publishes a secure software development framework and a map of attacks in NIST AI 1002 Edition 2023 that classifies availability, integrity, privacy and misuse risks, and promotes the risk budget approach. Warns that guardrails are theoretically brittle, supply‑chain backdoors can invisibly inject vulnerable code into coding assistants, and organizations must monitor runtime, track data provenance, and plan recovery.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 0:26Falling Water AI metaphor
- 3:19Trustworthy AI attributes
- 5:24Tradeoffs in trustworthy AI
- 7:10AI risks and NIST standards
- 8:54Adversarial AI attacks
- 11:25AI lifecycle risks
- 12:07Risk budget for AI security
- 13:12AI risk management
- 14:43Generative AI readiness
- 15:47AI security perimeter
- 17:53AI supply chain issues
- 19:59Q&A
related talks
35mAI and IoT security - The road ahead. Expert Panel
31mSecuring the Future: Navigating AI Vulnerabilities and Evolving Security Practices
49mThe Intersection of AI and Cybersecurity
14mDesigning the open and safe AI future
27mThinking Ahead About AI Security and Privacy
49mGenAI and Cybersecurity: Safeguarding the Digital Age
37mFundamentals of AI safety and security
05mThe Asymmetry of AI Security and Physical Hardware Resilience