https://trydoberman.dev/blog/ Field notes from building Doberman: AI agent security, runtime authorization, and lessons from real incidents. en https://trydoberman.dev/blog/approval-fatigue-is-a-security-issue/ https://trydoberman.dev/blog/approval-fatigue-is-a-security-issue/ Fri, 28 Aug 2026 17:02:19 +0000 I’ve spent the last two weeks making Doberman ask for approval less often. This increased safety and here’s why. https://trydoberman.dev/blog/cisco-defenseclaw-vs-doberman/ https://trydoberman.dev/blog/cisco-defenseclaw-vs-doberman/ Sat, 08 Aug 2026 23:34:11 +0000 Cisco's answer to runtime AI defense — what DefenseClaw gets right, and where a dedicated authorization layer makes a different bet. https://trydoberman.dev/blog/ai-security-playbook/ https://trydoberman.dev/blog/ai-security-playbook/ Thu, 06 Aug 2026 15:11:32 +0000 Takeaways from RAND's Practical Guide for Securing AI Models — and why AI security is shifting from model safety to execution security. https://trydoberman.dev/blog/when-ai-output-stops-being-text/ https://trydoberman.dev/blog/when-ai-output-stops-being-text/ Mon, 03 Aug 2026 01:26:29 +0000 The Hugging Face incident shows why agent security cannot end with better prompts, safer models or stronger sandboxes.