AI in Security: Navigating Ownership and Control скачать в хорошем качестве

AI in Security: Navigating Ownership and Control

If your AI can find a vulnerability, draft a patch, and open a pull request, your biggest risk is no longer detection quality. It is workflow ownership. This episode turns this week's security and governance signals into a control contract operators can run immediately. In this episode: Security workflow ownership from analyze to deploy Approval thresholds for AI-assisted remediation Evidence and logging standards for incident/legal readiness Monday control block you can run in 45 minutes Episode page: https://www.michaelhbm.com/AIChangeDe... RSS episode: https://rss.com/podcasts/aichangedesk... Apple Podcasts: https://podcasts.apple.com/us/podcast... Spotify: https://open.spotify.com/show/5X1sLLT... Sources: https://openai.com/index/codex-securi... https://www.axios.com/2026/03/06/open... https://www.anthropic.com/news/mozill... https://blog.mozilla.org/en/firefox/h... https://www.anthropic.com/coordinated... https://www.nist.gov/publications/cha... https://www.nist.gov/caisi/ai-agent-s... https://openai.com/index/introducing-... https://openai.com/index/chatgpt-for-... Production disclosure: AI-assisted tooling was used in parts of research and production support. Final editorial decisions and release approval remained human-led. Chapters: 00:00 Cold open + show contract 02:10 Why this episode now (EP005 + EP006 bridge) 04:20 Story 1: Codex Security and workflow ownership 08:20 Story 2: Anthropic + Mozilla CVD workflow lanes 12:45 Story 3: NIST AI 800-4 monitoring control pack 16:40 Story 4: GPT-5.4 + Excel execution-surface shift 19:50 Failure scenario replay with control layering 22:05 Monday action block (45 minutes, one owner) 23:40 Resistance handling + scorecard 24:35 30-60-90 path, close, and outro