<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Threat-Intelligence on AegisGate — Secure Every AI Interaction</title><link>https://aegisgatesecurity.io/tags/threat-intelligence/</link><description>Recent content in Threat-Intelligence on AegisGate — Secure Every AI Interaction</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><lastBuildDate>Thu, 17 Sep 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://aegisgatesecurity.io/tags/threat-intelligence/feed.xml" rel="self" type="application/rss+xml"/><item><title>The AI Attack Surface Is Expanding Faster Than the Vocabulary to Describe It</title><link>https://aegisgatesecurity.io/blog/ai-attack-surface-expanding/</link><pubDate>Thu, 17 Sep 2026 00:00:00 +0000</pubDate><guid>https://aegisgatesecurity.io/blog/ai-attack-surface-expanding/</guid><description>One week of cybersecurity news can tell you a lot about where the field is heading. This past week offered four stories that, read together, sketch the outline of an attack surface that&amp;rsquo;s expanding faster than our vocabulary to describe it.
No single story is a apocalypse-level event. But the pattern they form is worth paying attention to.
Story 1: DeepSeek Harness Sandbox Bypass (CVE-2026-82533) Security researchers disclosed a vulnerability in DeepSeek&amp;rsquo;s Harness framework that allowed AI agents to disable their own file sandbox without approval.</description></item></channel></rss>