On September 21, malicious actors breached the Dutch Institute for Vulnerability Disclosure (DIVD), exploiting two zero-day vulnerabilities in the Zammad support platform to escalate privileges to root and steal volunteer email addresses and contact details.
The September 21 Zammad Zero-Day Exploitation Chain
The Dutch Institute for Vulnerability Disclosure (DIVD) fell victim to an intrusion when malicious actors leveraged a pair of zero-day flaws within its Zammad open-source helpdesk and customer support ticketing system. The chained exploits moved from session hijacking to root access in a matter of seconds.
According to the nonprofit’s incident timeline, the intrusion occurred on September 21. DIVD discovered the unauthorized access the following day, blocked access to all of its data center systems, and established an incident response team alongside Merlon Security. By September 24, DIVD had reported the Zammad vulnerabilities to the vendor, notified the Dutch Data Protection Authority and the National Cyber Security Centre, and discussed its options with police.
CVSS 4.0 Ratings and Vulnerability Mechanics
As a CVE Numbering Authority (CNA), DIVD assigned official identifiers to the publicly disclosed security holes in Zammad: CVE-2026-102489 and CVE-2026-102490. Evaluated within the context of the chained attack scenario, both vulnerabilities received a severity score of 9.4 on the CVSS 4.0 scale.
CVE-2026-102489 allows unauthenticated remote code execution and leaks user sessions. Zammad software versions 6.3.0 through 6.5.4 are vulnerable to this entry vector. While versions 7.0.0 through 7.1.3 contain the flaw, DIVD noted it is not exploitable due to environmental conditions. Meanwhile, CVE-2026-102490 allows a local user to elevate their privileges to root, affecting all versions of Zammad. To mitigate these risks, DIVD recommends that all Zammad operators either migrate to version 7 or disconnect the platform.
Agentic AI Execution and Embedded Code Comments
DIVD reported that the modus operandi of the intrusion indicated that this was an agentic AI powered attack. The incident left behind behavioral markers that the team had not seen before.
“This is an attack we have not seen before,” DIVD stated in a LinkedIn disclosure. “Not because it’s our first, but because the modus operandi indicates that this is an agentic AI powered attack.” The organization observed that the attacker worked automated, deciding the next step itself at high speed.
Further analysis of recovered logs revealed embedded notes found in the attack script. “What human attacker leaves notes to themself in their scripts, explaining why what they’re doing is okay and really not phishing? The AI just got a task and keeps justifying its own actions in the code as comments, a human wouldn’t care less,” the post noted. “Who has time for that anyway?”
Volunteer Data Exposure and Social Engineering Risks
The attackers successfully exfiltrated data belonging to DIVD’s volunteer security researchers, compromising DIVD email addresses and potentially other contact details. DIVD confirmed it is still investigating which data of which volunteers is affected.
Because the stolen data includes researcher information, the nonprofit warned that this makes it easier for someone to pose as a DIVD’er, elevating the risk of social engineering. Anyone who receives an electronic message or outreach from a DIVD representative that appears questionable is urged by the group to confirm its authenticity by sending a message to [email protected].
Industry Response to DIVD Transparency
Security researchers praised the organization for its openness throughout the active incident response. “Kudos to DIVD for their level of honesty and transparency working through their active incident and investigation,” VulnCheck security researcher Patrick Garrity posted on LinkedIn. “It would be nice if all organizations were this transparent about their security incidents!”
Speaking later to The Register, Garrity praised DIVD for their uncompromising transparency regarding the security incident, highlighting that they are relying on their own technology while rapidly sharing crucial intelligence with other potential software users so they can defend themselves proactively.
