<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Adversarial Testing on Truth-First Beacon — Paul Desai</title><link>https://beacon.activemirror.ai/tags/adversarial-testing/</link><description>Recent content in Adversarial Testing on Truth-First Beacon — Paul Desai</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Fri, 20 Feb 2026 18:02:28 +0530</lastBuildDate><atom:link href="https://beacon.activemirror.ai/tags/adversarial-testing/feed.xml" rel="self" type="application/rss+xml"/><item><title>Governance That Runs</title><link>https://beacon.activemirror.ai/reflections/governance-that-runs/</link><pubDate>Fri, 20 Feb 2026 18:02:28 +0530</pubDate><guid>https://beacon.activemirror.ai/reflections/governance-that-runs/</guid><description>&lt;p&gt;Governance becomes real when it enforces itself at runtime, not when you write it in a document.&lt;/p&gt;
&lt;p&gt;I built &lt;code&gt;governance_runtime.py&lt;/code&gt; because I got tired of aspirational sovereignty. Every system claims to respect privacy, conserve resources, maintain autonomy. Few of them actually enforce these constraints when the model is running. The gap between policy and execution is where most AI governance dies — not from malice, but from the simple fact that checking compliance is someone else&amp;rsquo;s problem.&lt;/p&gt;</description></item></channel></rss>