<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Islam Mesabah</title><description>Notes on LLMs, document AI and teaching.</description><link>https://islammesabah.github.io/</link><item><title>Try Privacy Guardrail: keep personal data out of your AI chats</title><link>https://islammesabah.github.io/blog/privacy-guardrail/</link><guid isPermaLink="true">https://islammesabah.github.io/blog/privacy-guardrail/</guid><description>A free, open-source Chrome extension from our team at DFKI that catches personal data before you paste it into ChatGPT, Claude or Gemini. Please try it, rate it and tell us what breaks.</description><pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate></item><item><title>Protecting LLM systems from prompt attacks</title><link>https://islammesabah.github.io/blog/protecting-llm-systems-from-prompt-attacks/</link><guid isPermaLink="true">https://islammesabah.github.io/blog/protecting-llm-systems-from-prompt-attacks/</guid><description>My article in Hakin9 Magazine&apos;s issue &quot;LLM Mayhem – Hackers&apos; New Anthem&quot; on the main types of prompt attacks and how to defend against them at the model, prompt and system level.</description><pubDate>Thu, 13 Feb 2025 00:00:00 GMT</pubDate></item></channel></rss>