<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Arctovec</title><description>In-depth writing on agentic AI, security, and the parts of software that quietly hold up the rest.</description><link>https://arctovec.tech/</link><language>en-us</language><copyright>© 2026 Tom Reinholt</copyright><atom:link href="https://arctovec.tech/rss.xml" rel="self" type="application/rss+xml"/><item><title>Shadow AI is your governance problem now</title><link>https://arctovec.tech/posts/shadow-ai-governance-problem/</link><guid isPermaLink="true">https://arctovec.tech/posts/shadow-ai-governance-problem/</guid><description>Why blanket ChatGPT bans don&apos;t work, what the data on employee AI use actually shows, and the governance shape that survives the next two years.</description><pubDate>Fri, 22 May 2026 00:00:00 GMT</pubDate><category>shadow-ai</category><category>shadow-ai</category><category>governance</category><category>enterprise</category><category>data-loss-prevention</category><author>tom@arctovec.tech (Tom Reinholt)</author></item><item><title>AI agents are the new cybersecurity nightmare</title><link>https://arctovec.tech/posts/ai-agents-cybersecurity-nightmare/</link><guid isPermaLink="true">https://arctovec.tech/posts/ai-agents-cybersecurity-nightmare/</guid><description>Why the Slack agent that rewrote its own policy isn&apos;t an outlier — and what RSAC quietly admitted about the next twelve months.</description><pubDate>Wed, 20 May 2026 00:00:00 GMT</pubDate><category>agents-vs-security</category><category>agents</category><category>rsac</category><category>agentic-ai</category><author>tom@arctovec.tech (Tom Reinholt)</author></item></channel></rss>