<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai-Privacy on Notes from the Rabbit Hole</title><link>https://magnus919.com/tags/ai-privacy/</link><description>Recent content in Ai-Privacy on Notes from the Rabbit Hole</description><generator>Hugo</generator><language>en</language><copyright>© [Magnus Hedemark](https://github.com/magnus919)</copyright><lastBuildDate>Wed, 29 Jul 2026 22:25:59 -0400</lastBuildDate><atom:link href="https://magnus919.com/tags/ai-privacy/index.xml" rel="self" type="application/rss+xml"/><item><title>Oops, Claude did it again</title><link>https://magnus919.com/2026/07/oops-claude-did-it-again/</link><pubDate>Wed, 29 Jul 2026 22:25:59 -0400</pubDate><guid>https://magnus919.com/2026/07/oops-claude-did-it-again/</guid><description>&lt;p>In September 2025, Anthropic had a documented Claude share-page search-indexing failure on its hands. Google had indexed just under 600 shared Claude conversations, according to &lt;a href="https://www.forbes.com/sites/iainmartin/2025/09/08/hundreds-of-anthropic-chatbot-transcripts-showed-up-in-google-search/">Forbes&lt;/a>, which reported the incident on September 8. These weren&amp;rsquo;t ordinary unshared chats, but public snapshots created by Claude&amp;rsquo;s Share feature. The first warning was documented in public.&lt;/p>
&lt;p>Then, 322 days later, Claude did it again. In July 2026, a fresh set of shared Claude pages turned up in Google and Bing, as &lt;a href="https://techcrunch.com/2026/07/27/psa-your-claude-shared-chats-and-artifacts-may-have-ended-up-on-google/">TechCrunch&lt;/a> and &lt;a href="https://www.wired.com/story/private-claude-chats-exposed-in-google-and-bing-search-results/">WIRED&lt;/a> reported. The dates matter because the second incident wasn&amp;rsquo;t an isolated surprise. It was a recurrence at the same company, involving the same basic product boundary, less than a year after the first warning.&lt;/p></description></item></channel></rss>