The short version
- The disclosure. Abdelhamid Naceri, who publishes as Nightmare Eclipse and MSNightmare, released SignDrop proof-of-concept code and a writeup on GitHub in the same week Microsoft shipped its November security updates. He estimates that 80% of Windows installations are vulnerable.
- How it works. A standard user prepares a Common Log File System (CLFS) log and container with attacker-chosen paths, then coaxes Windows Defender into opening the log. The driver creates the .cnpf authentication file with SYSTEM privileges but the original file's permissions, which can be chained into a SYSTEM shell.
- No fix yet. Microsoft has not confirmed the flaw, assigned a CVE or patched it in the November updates. The exploit is not in CISA's catalogue of vulnerabilities known to be exploited, and no attacks using it have been reported.
- Independent testing. Korean firm 78researchlab says its researcher Heewon Song tested the code and found it works but is not reliable enough for operational use. Cyderes compiled and tested the proof of concept and saw no evidence of exploitation in the wild.
- The workaround is worse. Turning off CLFS logfile authentication takes a machine outside the reported attack surface, but the researcher and Microsoft's own documentation warn that it lets the driver parse tampered logs again, reopening the memory-corruption flaws the check was added to stop.
Timeline
| Date | Event |
|---|---|
| October 2025 | Microsoft adds logfile authentication to the Common Log File System driver in updates for Windows 11 25H2, 26H1 and Windows Server 2025. |
| November | Naceri publishes the SignDrop proof-of-concept code and a detailed writeup on GitHub on a Tuesday, the same day Microsoft releases its November security updates. |
| 12 Nov | The exploit is reported publicly; 78researchlab publishes an assessment saying the code works but does not run reliably enough for operational use. |
| 13 Nov | SecurityWeek reports that Cyderes compiled and tested the proof of concept and found no evidence the flaw has been exploited in the wild. |
What SignDrop does
CLFS is a logging subsystem in Windows that parses log files in kernel mode. Its driver has long been a target: 10 CLFS flaws sit in CISA's catalogue of vulnerabilities known to be exploited, including one used in ransomware campaigns in April 2025. In October 2025 Microsoft added logfile authentication, which signs log data with a key unique to each machine so the driver only parses logs it wrote; a companion .cnpf file holds recovery data.
SignDrop attacks the creation of that companion file. A standard user prepares a log and container with paths they control. When a privileged process opens the log — in the published code, the Defender antimalware service, triggered through its signature update interface — the driver creates the .cnpf file in that process's security context. Symbolic links redirect the write to a chosen location, and the file inherits the original log's permissions, leaving a SYSTEM-created file the standard user can write to.
Naceri says Microsoft built the mitigation correctly but that "the logic allows abuse". He argues the same approach could be used against any antimalware engine running with elevated privileges, because the file-creation request appears to come from the engine's own process. That broader claim has not been tested, and the published chain requires Defender to be enabled.
Who is exposed, and what is confirmed
Naceri says he tested the exploit on Windows 11 26H2, a Windows 11 26H2 Experimental build and Windows Server 2025 Experimental. He says installations older than 90 days are generally vulnerable, because they have moved from learning to enforcement mode. Administrators can read the Mode value under HKLM\SYSTEM\CurrentControlSet\Services\CLFS\Authentication: 0 or 1 falls inside the claimed scope, while 2 or 3 means the authentication is switched off. The sources differ on which of 0 and 1 is learning and which is enforcement.
Microsoft has not confirmed the vulnerability, assigned a CVE or released a fix, and nothing for SignDrop appeared in the November security updates. The exploit is not in CISA's Known Exploited Vulnerabilities catalogue and no attacks using it have been reported. 78researchlab's head of research, Sungyup Nam, says researcher Heewon Song tested the code and found it works, but judged it too unreliable for operational use. Cyderes also compiled and tested it.
The published attack is loud. Running it changes every UAC prompt to open a standard user shell and a SYSTEM shell, blocking other actions, so anyone at the machine would notice. Naceri says the only remedy is to wait for a patch, since disabling the CLFS mitigation would expose devices to the memory-corruption attacks it was introduced to stop.
A long run of Windows zero-days
Naceri, who says he is a former Microsoft employee, has disclosed a series of Windows and Defender flaws since April, including BigDiskBuster, ShieldCrash, ShieldBreak, LegacyHive, RoguePlanet, BlueHammer, RedSun, YellowKey, GreenPlasma, MiniPlasma and UnDefend. Microsoft confirmed RoguePlanet in June as CVE-2026-50656 and patched it the following month; ShieldBreak was assigned CVE-2026-69414 in August. According to The Hacker News, three of his earlier Defender flaws have been exploited in attacks.
Those earlier issues were race conditions in Defender's file handling. SignDrop is different: it abuses the CLFS driver's own authentication feature, added to reduce exposure to CLFS memory-corruption attacks. After earlier disclosures made without private reporting, Microsoft warned of legal action against anyone causing real harm to its customers, which many in the industry read as a threat to the researcher.
What is not known yet
- Whether Microsoft will confirm SignDrop as a vulnerability, assign a CVE and patch it, and on what timeline.
- Whether the technique works against antimalware engines other than Defender, as the researcher claims; only the Defender chain has been tested.
- How many systems are actually vulnerable; the 80% figure is the researcher's estimate rather than a measurement.
- Whether anyone has used the exploit outside testing; Cyderes found no evidence, but there is no broad telemetry.
- Whether the exploit can be made reliable enough for real attacks, given the loud side effects reported by testers.
What to do
If you administer Windows systems
- Read the Mode value under HKLM\SYSTEM\CurrentControlSet\Services\CLFS\Authentication; a value of 0 or 1 puts the machine inside the reported scope.
- Do not set that value to 2 or 3 to sidestep the issue; it disables logfile authentication and restores the older memory-corruption exposure.
- Watch Microsoft's security update guide for an advisory or fix before changing anything else.
If you plan to validate exposure
- Use the registry check or the researcher's proof of concept on a test machine, not on production systems, because the published chain alters UAC prompts and can block other actions.
- Note that the published chain needs Defender enabled, so a successful test tells you about that configuration rather than about every Windows build.
For everyone else
- Treat SignDrop as a local privilege escalation: an attacker first needs a way to run code on the machine before it can be used.
- Keep security features enabled even when they prove abusable; disabling this one brings back the exploited flaws it was built to block.
Update log
- First published.
Sources
New incident reports by email
One email a week with the reports published since the last one.