That doesn't mean every Windows PC can be taken over remotely. SignDrop is a local privilege escalation bug, so an attacker needs some other way onto the machine first. Once a phishing payload or a compromised standard account gets them there, though, this kind of bug is what turns a limited foothold into full control.
What was disclosed, and when
SecurityWeek reported that security researcher Abdelhamid Naceri, known online as Nightmare Eclipse, this week dropped another zero-day exploit targeting Windows systems, and that the exploit allows a local user to elevate their privileges to System by abusing flawed logic in the Windows Common Log File System (CLFS).
Some early summaries gave the release date as November 17. Security reporting puts it a week earlier. ThreatDefence wrote that on November 10, 2026, a researcher publishing as MSNightmare on GitHub released SignDrop, a local privilege escalation zero-day in the Windows Common Log File System (CLFS) driver, clfs.sys, with working proof-of-concept code. SecurityWeek's story ran November 13. Wisevoter's later summary, which credits RocketNews, also reports that researcher Will Dormann confirmed the exploit works as described. That confirmation came to us second-hand and we could not check it against Dormann's own statement.
Key facts:
- Type: Local privilege escalation from a standard user to SYSTEM
- Components involved: The CLFS driver (clfs.sys) and Microsoft Defender's signature-update interface
- Public exploit code: Yes, a proof of concept on GitHub
- CVE or patch: None reported so far
- Exploitation in the wild: None observed, according to Microsoft
A security fix used against itself
SignDrop targets a hardening feature, not an old bug. Microsoft added it after attackers spent years abusing CLFS. CLFS parses complex log files in kernel mode, which made it a popular target for privilege escalation.
According to Microsoft's support documentation (KB5056852), the mitigation adds a hash-based message authentication code (HMAC) to the files behind each CLFS logfile. The key used to create those codes is unique to each system, is stored in the registry, and can only be read by Administrators and SYSTEM. When enforcement is on, CLFS refuses to open a logfile whose authentication fails. The feature shipped in Windows 11 version 25H2 and Windows Server 2025 updates released on or after October 28, 2025. Microsoft's page also lists Windows 11 version 26H1.
The feature creates a companion "patch file" next to each logfile. Microsoft gives this example: a logfile at C:\Users\User\example.blf gets a patch file at C:\Users\User\example.blf.cnpf. SignDrop targets the code that creates that .cnpf file.
How the attack chain reportedly works
The details below come from the researcher and from people who have studied the proof of concept. Microsoft has not published its own root-cause analysis.
- The user prepares a malicious log. According to the proof-of-concept code, a standard user prepares a CLFS log and container with paths the user controls.
- Defender is made to open it. A standard user can create a log file and specify a path to an associated container in it, and then coerce Windows Defender to open the file in the context of the antimalware engine by calling the signature update RPC interface.
- CLFS writes the signature file with SYSTEM privileges. SecurityWeek reports Naceri's explanation that SignDrop exploits a vulnerability in how clfs.sys handles the creation of the authentication file '.cnpf' during the creation of a log file. If the signature file doesn't exist yet, it gets created with System privileges at a location the attacker influenced.
- The payload is loaded. SecurityWeek says the proof of concept then reopens that file and puts a payload in it, which ends with a System shell. Reports also say Windows Object Manager symbolic links are part of the technique.
Each step uses legitimate behavior: a logging driver signing its own files and an antivirus engine opening files during an update. The trick is in how the steps are combined. The Hacker News also points out that where the earlier exploits targeted race conditions in Defender's file handling, this one targets the CLFS driver's own authentication feature.
Who is actually exposed?
The submitted story ties exposure partly to how long a system has been running. That comes from the researcher's rule of thumb. SecurityWeek quoted Naceri saying that "You're generally vulnerable if your Windows installation is older than 90 days", and he also estimated SignDrop would work on about 80% of Windows installations. Nobody has independently checked that 80% figure.
The more accurate test is the CLFS authentication mode:
Registry Mode value | Meaning (per Microsoft) | Within SignDrop's claimed scope? |
|---|---|---|
| 0 | Enforcement: CLFS rejects logfiles with missing or invalid codes | Yes (researcher/Cyderes claim) |
| 1 | Learning: CLFS adds missing codes when it opens a logfile | Yes (researcher/Cyderes claim) |
| 2 | Disabled by an administrator | Outside claimed scope |
| 3 | Disabled automatically by the system | Outside claimed scope |
The 90-day figure comes from how the feature rolls out. Microsoft says learning mode currently lasts 90 days from the first boot with the new CLFS driver. After that, the driver switches to enforcement mode the next time it starts. Because learning mode (Mode 1) is also listed as vulnerable, a machine that is only a week into its adoption period may still be in scope. System age alone doesn't tell you much.
Two more conditions are reported:
- Tested builds: The exploit was tested on Windows 11 26H2, a Windows 11 26H2 Experimental build, and Windows Server 2025 Experimental, according to the repository. Other releases that ship the feature, such as 25H2, are inferred to be in scope from the mode condition. They weren't listed as tested.
- Defender must be running: According to The Hacker News, Defender must be enabled for the published chain to succeed.
Microsoft's response so far
Microsoft gave SecurityWeek a fairly specific statement. It said it is actively investigating, that Defender detects and blocks exploitation and alerts on the related privilege-escalation behavior, and that it has not seen exploitation in the wild. Cyderes says it compiled and tested the proof of concept and also found no evidence of in-the-wild use.
So the Defender service that the exploit relies on is also, according to Microsoft, what detects it. That helps only if Defender's signatures are current, and only on machines where Defender is the active antimalware engine. Microsoft's statement does not say a code fix has shipped. The Hacker News reported that no CVE, advisory, or patch for SignDrop appeared in the November security updates.
What administrators should do now
1. Check the CLFS authentication mode (read-only)
Microsoft documents the setting at HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\CLFS\Authentication, in the Mode value. From an elevated PowerShell prompt you can read it without changing anything:
Get-ItemProperty 'HKLM:\SYSTEM\CurrentControlSet\Services\CLFS\Authentication'
A value of 0 or 1 puts the machine inside the researcher's claimed scope. One security write-up notes that builds without CLFS authentication don't have the Authentication key. This check only tells you whether the conditions are met. It doesn't prove the machine can be exploited.
2. Do not disable CLFS authentication as a "fix"
Turning the mitigation off can look like the fast fix. Microsoft's documentation says administrators can disable it by setting Mode to 2, either directly or with the "Enable / disable CLFS logfile authentication" policy under Computer Configuration > Administrative Templates > System > Filesystem. Microsoft doesn't recommend it, and it says admins should never set the value to 3 by hand. Disabling the feature would also bring back the older CLFS memory-corruption attacks it was built to stop. Microsoft has not said that disabling it blocks SignDrop.
3. Keep Defender current and enabled
Microsoft's detection claim depends on Defender signatures being up to date. Keep installing Windows and Defender updates while you wait for a code fix, and confirm that tamper protection and real-time protection haven't been switched off on your endpoints.
4. Hunt for the public PoC's footprints
Defenders have started publishing detection ideas. ThreatDefence published sample Defender XDR queries that look for the file names the public project uses, including hid.dll loaded from outside the normal System32, SysWOW64 and WinSxS folders. Secure.com lists unusual .cnpf and .blf files and SYSTEM-level conhost.exe activity tied to MsMpEng.exe as indicators. These patterns match the public code, so an attacker who modifies it can avoid them. Use them as a starting point and keep looking beyond them.
5. Reduce standard-user code execution
The attacker needs local code execution first, so the usual controls still help: application control (WDAC/App Control or AppLocker), limiting who can log on interactively to servers, and fast containment of phishing footholds.
The bigger picture
CLFS has been a favorite privilege-escalation target for years, and logfile authentication was supposed to close off that whole class of attacks. SignDrop doesn't defeat the HMAC cryptography. It abuses the file-creation logic that runs alongside it, and it uses a privileged process (Defender) that will open files when asked. New hardening features add new code that runs with high privileges, and that code can be attacked too.
The disclosure also continues the debate over full public release. Naceri has published several Windows zero-days with working exploit code, so defenders and attackers get the code at the same moment. Supporters say this forces vendors to act. Critics say it leaves organizations exposed with no patch. Either way, the practical position for Windows admins is the same: there's no patch and no CVE yet, but Defender detection exists and Microsoft is investigating.
Bottom line
- SignDrop is a public, unpatched local privilege escalation exploit that turns a standard user into SYSTEM through CLFS logfile authentication and Defender's signature-update RPC interface.
- Exposure depends on the CLFS authentication
Mode(0 or 1, per the researcher and Cyderes). Machine age is a rough proxy at best. - Microsoft is investigating, says Defender detects and blocks the exploit, and reports no in-the-wild use.
- Check the registry value, keep Defender and Windows updated, and leave the mitigation turned on. Install Microsoft's fix as soon as it ships, whether as a Patch Tuesday update or a Defender platform update.
References
- Researcher Discloses SignDrop Windows Privilege Escalation Flaw Wisevoter · 2026-11-20T22:43:37+00:00
- Windows SignDrop CLFS Zero-Day: Mitigation and Hunting threatdefence.com
- Nightmare Eclipse Drops Windows Zero-Day Exploit 'SignDrop' - SecurityWeek securityweek.com