<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>AI Security on The Broomstick Brief</title><link>https://broomstick.tymyrddin.dev/tags/ai-security/</link><description>Recent content in AI Security on The Broomstick Brief</description><generator>Hugo -- 0.147.3</generator><language>en</language><lastBuildDate>Sat, 25 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://broomstick.tymyrddin.dev/tags/ai-security/index.xml" rel="self" type="application/rss+xml"/><item><title>The Glasswing problem</title><link>https://broomstick.tymyrddin.dev/posts/glasswing/</link><pubDate>Sat, 25 Apr 2026 00:00:00 +0000</pubDate><guid>https://broomstick.tymyrddin.dev/posts/glasswing/</guid><description>The framing was unmistakable. We have built something dangerous. We are choosing not to release it. Be grateful that we are the ones who built it. Why both halves of that framing deserve more scrutiny than they have received, and why the genuinely interesting story, the structural one, has been almost entirely buried under coverage of the headline numbers.</description></item></channel></rss>