The first post-Mythos Patch Tuesday was a big one: Microsoft pushed out fixes for 164 vulnerabilities; eight critical and two actively exploited.
Including third-party and Chromium fixes bundled into Redmond's monthly security cycle, today's update amounts to 247 security patches.
How many were found using AI? Hard data would be hard to gather this early; anecdote isn’t. Here’s one: security researcher Joe Desimone was credited by Microsoft with five local privilege escalation (LPE) bugs and noted pointedly on X that “all [were] found with AI (not Mythos)”.
Two days, nine additions to KEV...
The two known-exploited included in April’s Patch Tuesday were: CVE-2026-32201, a Microsoft SharePoint server spoofing bug, and CVE-2026-5281, a use-after-free bug in Chromium, used by Edge.
They join the seven vulnerabilities that hit CISA’s KEV on Monday: (CVE-2012-1854, CVE-2020-9715, CVE-2023-21529, CVE-2023-36424, CVE-2025-60710, CVE-2026-34621 , and CVE-2026-21643 – yes, the latter is a SQL injection vulnerability, in Fortinet’s FortiClient EMS, in 2026.*
Many security leaders believe that AI is about to ramp up the volume of exploited zero days at significant pace. They wrote in a “Mythos Ready” paper published on April 14 that organisations should be ready for a cadence and volume of vulnerability disclosures [that] will exceed anything we have experienced before” and to “re-orient security program resources for… decreasing time to patch, and more persistent and complex attacks.”
"Build the muscle now"
“The organizations that respond well will be those that build the muscle now: the processes, the tooling, and a culture willing to adopt AI as a core part of how security gets done,” its authors wrote – admitting that for now “most AI defensive controls and approaches are not yet mature…”
As ever, patch up as fast as you can.
As the Mythos Reader authors emphasised: "The basics remain valid and can be prioritized for risks that can’t be easily mitigated. Implement egress filtering (it blocked every public log4j exploit). Enforce deep segmentation and zero trust where possible. Lock down your dependency chain.
"Mandate phishing-resistant MFA for all privileged accounts.
"Every boundary increases attacker cost.
"... software minimization is a high leverage function that reduces the operational overhead of second order functions such as patching. For example, minimizing base operating system images, or replacing third-party libraries with framework primitives as they emerge over time. AI can do this."
*It’s also the second zero day exploited in FortiClient this month alone, after CVE-2026-35616 hit headlines. The latest FortiF*ckup was introduced in version 7.4.4, with a refactoring introducing the bug which means an HTTP header used to identify which tenant a request belongs to is passed directly into a database query without sanitisation before any login check.
Per Bishop Fox’s short analysis here: “An attacker who can reach the EMS web interface over HTTPS [yes, this shouldn’t happen, but does] needs no credentials to exploit this. A single HTTP request with a crafted header value is sufficient to execute arbitrary SQL against the backing PostgreSQL database. This gives attackers access to admin credentials, endpoint inventory data, security policies, and certificates for managed endpoints.”
The link has been copied!
