<?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>Anthropic on jamesm.blog</title>
    <link>https://jamesm.blog/tags/anthropic/</link>
    <description>Recent content in Anthropic on jamesm.blog</description>
    <image>
      <title>jamesm.blog</title>
      <url>https://jamesm.blog/papermod-cover.png</url>
      <link>https://jamesm.blog/papermod-cover.png</link>
    </image>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Wed, 08 Apr 2026 08:45:00 +0100</lastBuildDate>
    <atom:link href="https://jamesm.blog/tags/anthropic/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Claude Mythos: The AI Benchmark Breaker That Won&#39;t Be Released</title>
      <link>https://jamesm.blog/ai/claude-mythos-benchmarks/</link>
      <pubDate>Wed, 08 Apr 2026 08:45:00 +0100</pubDate>
      <guid>https://jamesm.blog/ai/claude-mythos-benchmarks/</guid>
      <description>Claude Mythos Preview shattered AI benchmarks with 93.9% SWE-bench and 97.6% USAMO scores, but remains restricted to security partners due to cybersecurity risks.</description>
    </item>
    <item>
      <title>Claude Code Source Leak: Anthropic&#39;s 2,000-File Exposure and What It Means</title>
      <link>https://jamesm.blog/ai/claude-code-source-leak/</link>
      <pubDate>Wed, 01 Apr 2026 07:42:00 +0000</pubDate>
      <guid>https://jamesm.blog/ai/claude-code-source-leak/</guid>
      <description>Nearly 2,000 internal files from Claude Code were briefly leaked due to human error. Here&amp;#39;s what happened, what was exposed, and the security implications for developers using Anthropic&amp;#39;s AI coding tool.</description>
    </item>
  </channel>
</rss>
