# SignDrop: Windows zero-day turns Defender against itself

> SignDrop, a public Windows zero-day, abuses Microsoft's CLFS mitigation and Defender to reach SYSTEM. No patch exists; a registry check shows exposure.

- Published: 2026-11-13
- Topic: [Security](https://futuretechnologyhq.com/topic/security/)
- URL: https://futuretechnologyhq.com/article/signdrop-windows-zero-day-turns-defender-against-itself/
- Publisher: [Future Technology HQ](https://futuretechnologyhq.com/), free to read. Quote as "Future Technology HQ" with a link to the article.

## Key points

- SignDrop gives a standard local user SYSTEM privileges by tricking Windows Defender into creating a CLFS authentication file
- Check HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\CLFS\Authentication\Mode: 0 or 1 means vulnerable, 2 or 3 means protected
- Disabling the CLFS mitigation removes the SignDrop path but reopens memory corruption exposure, so the only real fix is a Microsoft patch
- The same trick may let attackers disable other antivirus and EDR self-defence, not just Defender

## Article

A researcher using the name Chaotic Eclipse has published working proof-of-concept code for SignDrop, a Windows zero-day that lets a standard local user reach SYSTEM privileges. The biggest implication is that the flaw sits inside a Microsoft security mitigation, and the exploit can reportedly be pointed at other antivirus and EDR products to switch off their self-defence. There is no patch.

## What changed

The vulnerable component is the Common Log File System (CLFS), a Windows feature that parses complex log files in kernel mode. That design made it a repeat target for memory corruption bugs. Microsoft's answer was a CLFS authentication mitigation, rolled out to the public release channel from 28 October 2025. It works in two stages: 90 days in learning mode, then full protection, where CLFS requires every log file to carry a valid authentication code.

The researcher says SignDrop exists because of that staged rollout. His estimate is that about 80% of Windows installations worldwide are vulnerable, either because they are still inside the 90-day learning period or because the mitigation is only partly enabled. His own advice is blunt: any installation older than 90 days should be assumed vulnerable.

The mechanics are neat. The bug lives in clfs.sys, in how the driver creates the ".cnpf" authentication file tied to a new log container. When a process creates its own log files, CLFS applies proper access checks. The problem appears when a log file has an associated container. Opening the original file also opens the linked one, and if that linked file is empty, CLFS tries to sign it whether or not a signature file already exists. If a more privileged process does the opening, the signature file is created with that process's rights.

An attacker needs no special permission to open a CLFS log file, only the correct path. So a standard user prepares a file, then coerces Windows Defender into opening it by calling ServerMpUpdateEngineSignature and manipulating symbolic links in the Windows Object Manager. Defender follows the path stored in the original container, finds no signature, and creates one with SYSTEM privileges. CLFS then builds the new authentication file using the original file's DACL, the access control list that says who may touch it.

The researcher tested the code on Windows 11 26H2, Windows 11 26H2 Experimental and Windows Server 2025 Experimental. In his words, Microsoft did everything properly; "the logic allows abuse."

## Impact on businesses

The exposure is broad because the trigger condition is simply a Windows build older than 90 days. Any organisation running Windows endpoints or servers, in the UK or the US, should assume it is in scope until it has checked.

Local privilege escalation is not remote code execution, so an attacker first needs a foothold. That is exactly why these bugs matter in practice: ransomware crews and intruders routinely land on a machine as a low-privilege user and need a way up. A reliable public path to SYSTEM shortens that step.

The more serious claim is about security tooling. The researcher argues the same coercion works against any antimalware engine. Because the file creation appears to come from the antivirus vendor's own legitimate process, the vendor's kernel drivers let it through. His view is that this makes SignDrop usable to turn off any antivirus or EDR. That claim comes from the researcher, not from vendor confirmation, but it should worry SOC teams who treat tamper protection as a hard boundary.

For IT teams the practical test is simple. Check the registry key HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\CLFS\Authentication\Mode across the estate. A value of 0 or 1 means vulnerable; 2 or 3 means protected. The value is a single registry setting that can be read on each machine.

## Impact on consumers and users

The risk for home users is that malware already on a PC, perhaps a fake installer or a cracked app, can use SignDrop to gain SYSTEM and then undermine the antivirus meant to catch it. A machine older than 90 days since the mitigation reached it should be treated as exposed.

The sensible response is the usual one: install Windows updates promptly once a fix arrives, avoid running untrusted software, and keep backups that malware cannot reach. For the latter, see this guide to [immutable versus air-gapped backups](/article/immutable-vs-air-gapped-backups/).

## Impact on the wider industry

The awkward point for Microsoft is that, by the researcher's account, its mitigation for one bug class opened a new route to SYSTEM. Turning the mitigation off is no escape. The researcher notes that doing so "will expose your machine to memory corruption vulnerabilities as CLFS still does not handle log files well." Defenders face a trade between two risks until a patch ships.

For security vendors, the claim that self-defence can be bypassed across products is a pressure point. Chaotic Eclipse has form here. Besides Microsoft Defender zero-days such as Undefend, RedSun, BigDiskBuster and ShieldCrash, he has released exploits for Nvidia (GreenSection), Kaspersky Endpoint Security (HardBreacher), Avast (PrettyPrague) and CrowdStrike Falcon (FalconFlank), according to Security Affairs. Some of his earlier releases were later exploited in the wild. For a related Defender case, see our coverage of [ShieldBreak](/article/shieldbreak-defender-zero-day-cve-2026-69414/).

The disclosure argument is live too. The researcher publishes working code, often after criticising how vendors handle reports, and says he is Abdelhamid Naceri, a former Microsoft employee who describes a bitter dismissal dispute. He also admitted that an earlier claim that Microsoft had taken legal action against him was fabricated. Defenders will weigh his technical findings against the risks of publishing exploits before a fix exists.

## What comes next

Microsoft has not issued a fix in the material available. Security vendors will likely have to review how their self-defence drivers treat file creation from their own processes. Researchers and criminals can both study the public code in the meantime, and no patch has been announced. Microsoft's recent cadence is covered in our look at [Patch Tuesday](/article/microsoft-patch-tuesday-one-exploited-outlook-bug/).

## Key takeaways

- Check the CLFS Authentication Mode registry value now; 0 or 1 means exposed.
- Treat Windows installations older than 90 days as vulnerable.
- Do not disable the CLFS mitigation as a workaround without weighing the memory corruption risk it was built to stop.
- Do not assume EDR tamper protection holds against a local attacker until vendors respond.

## Verdict

Defenders get a public exploit, no patch and an uncomfortable choice between two risks. The researcher says the mitigation itself works as designed, but its logic can be abused. Watch for a Microsoft fix and for statements from antivirus vendors on whether their self-defence can be coerced the same way.

## Sources

- [Security Affairs](https://securityaffairs.com/202968/hacking/new-signdrop-exploit-can-abuse-windows-defender-to-bypass-security-protections.html)
- [securityweek.com](https://www.securityweek.com/nightmare-eclipse-drops-windows-zero-day-exploit-signdrop/)
