About
I'm Stefan Boesen, a security engineer focused on AI security and using AI for security.
This site is my primary home for longer-form writing on these topics and other technical subjects.
It's also a lightweight knowledge base for posts I want to keep organized and easy to revisit.
I take an AI-first approach to research, writing, and engineering on this site. That means treating models as part of the working toolchain rather than as a novelty: I use ChatGPT 5.4, ChatGPT 5.4 Pro, and Codex to accelerate research, pressure-test ideas, edit drafts, and help build the systems behind the writing.
This site is inspired by gwern.net and implements similar features like inline popins, reader mode, and dark mode.
For site controls and reading features, see the Help page.
For all posts, see the posts page.