<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Journal – Active Search</title><description>Articles from Active Search on AI software, AI safety and the EU AI Act.</description><link>https://activesearch.ai/</link><language>en</language><item><title>AI finds security flaws faster than they can be fixed</title><link>https://activesearch.ai/en/journal/safety/ai-finds-security-flaws/</link><guid isPermaLink="true">https://activesearch.ai/en/journal/safety/ai-finds-security-flaws/</guid><description>An AI model has found thousands of unknown vulnerabilities in widely used software. What Project Glasswing shows and what companies should take from it.</description><pubDate>Thu, 08 Oct 2026 00:00:00 GMT</pubDate><category>AI Safety</category></item><item><title>How strict should safeguards be?</title><link>https://activesearch.ai/en/journal/safety/how-strict-should-safeguards-be/</link><guid isPermaLink="true">https://activesearch.ai/en/journal/safety/how-strict-should-safeguards-be/</guid><description>AI providers restrict their models so they cannot be misused for attacks. That protects, but it also costs usefulness. Weighing it up.</description><pubDate>Thu, 08 Oct 2026 00:00:00 GMT</pubDate><category>AI Safety</category></item></channel></rss>