Last updated: 20 November 2026. This is a developing story based on a researcher's public claims and a handful of news reports. I could not test it: it needs a Windows 11 machine and, from what is reported, I would not run a public exploit anyway. Everything below is attributed to its source.
On 19 November 2026 a researcher known as Nightmare Eclipse (also Chaotic Eclipse), identified in news reports as Abdelhamid Naceri, published a proof of concept called FastUnlock. It claims that on a locked Windows 11 PC, someone with physical access can use the Narrator screen reader to reach settings and File Explorer that should be unavailable before sign-in. Headlines say it "bypasses BitLocker". The reporting does not support that claim as stated, and the difference decides whether you need to act this week. Here is what is known, what is not, and what to do on managed fleets and on your own laptop.
The short version
- Requirement: physical access, and, as reported, at least one account on the PC linked to a Microsoft account.
- Reported tested build: Windows 11 26H2 with the November 2026 updates. Other versions, including Windows 10, are untested by the researcher.
- Microsoft: as of 20 November, the coverage I read found no advisory, CVE or fix.
- What it is not (per one report): a break of BitLocker encryption. Cybersecurity News reports that it does not decrypt a powered-off drive. It gets past the sign-in screen on a machine that has already booted and unlocked its drive.
-
Independent verification: thin. ThreatLocker reportedly ran code as the lock screen's
WsiAccount. Other outlets say they could not reproduce it.
What is reported to happen
At a high level, according to Security Affairs and Cyber Kendra: the attack starts from the lock screen's account recovery flow, reached through the "forgot my PIN" or password option. Turning on Narrator there moves keyboard focus to a window behind the lock screen, and from Narrator's own settings the attacker reaches File Explorer and other settings. The researcher says this can then run a binary from a USB stick under WsiAccount, the account the lock screen uses, which is not an administrator. Cyber Kendra's explanation of the flaw is that a redesigned Narrator interface is not blocked from running at the lock screen, even though older code has that check. I am deliberately not reproducing the key sequences. They are in the news reports if you need them, and the defensive questions below do not depend on them.
The researcher's own claims go further: "about five minutes" to breach a machine, consistency "in every machine I tested", and a bypass of BitLocker on fully patched systems. The reports note that the public demo did not show code execution.
Does it bypass BitLocker?
This is where the headlines and the body text of the reports disagree. The researcher claims it defeats BitLocker. Cybersecurity News says the body of the story is that BitLocker encryption is not broken: the sign-in barrier is bypassed on a running, already unlocked system. It lists as not established that an attacker gets administrator rights, takes over the signed-in session, or can read every file.
Here is my own explanation of why both can be partly true; it is not from the researcher. BitLocker with a TPM only (no PIN) unlocks the drive automatically at boot, before you ever see the sign-in screen. The Windows lock screen is then the only thing standing between an attacker and the user's data on a machine that is already powered on or booted. A lock screen bypass therefore defeats the practical protection of a TPM-only laptop without touching the encryption. With TPM plus a startup PIN, the drive stays locked until the PIN is typed at boot, so a lock screen bug on a machine that is off, or shut down, does not get the attacker the data. That distinction matches what ThreatLocker said about the earlier YellowKey BitLocker bypass, which targeted TPM-only configurations and for which "a pre-boot PIN raises the bar". That article also reports that Microsoft recommended TPM+PIN.
The researcher has not shown, and the reports do not establish, that the Narrator bug works around a pre-boot PIN on a powered-off machine. Treat "BitLocker bypassed" as unproven until someone demonstrates it.
Is my machine affected?
On what is reported, you are in the risk group if all of these are true:
- It runs Windows 11 (tested on 26H2; others unknown).
- At least one user signs in with a Microsoft account, since the entry point is the recovery prompt for such accounts.
- Someone could get physical access: a laptop that travels, a kiosk, a shared desk, a lost or stolen device.
If nobody uses a Microsoft account on the device, the reported precondition is not met. A machine that has been fully shut down with a startup PIN is not exposed to the data-access scenario in the way a sleeping TPM-only laptop is, based on the reasoning above. I could not verify any of this on a real device.
What to do now
-
Find your TPM-only devices. Microsoft documents
manage-bde -statusand the PowerShell cmdletGet-BitLockerVolumefor listing protectors. I did not run them (no Windows machine here). A volume whose protectors list onlyTpmis TPM-only. -
Require a startup PIN where theft or loss is a real risk. The pre-boot PIN is the control ThreatLocker and Cybersecurity News both point to. It needs the Group Policy or Intune setting that requires additional authentication at startup, and then a
TpmAndPinprotector on the drive. Roll it out through policy rather than by hand, and plan the user and help-desk impact first: PIN resets and recovery key escrow need to work before you enforce it. - Prefer shutdown or hibernate over sleep for laptops in transit. Cybersecurity News notes that sleep keeps programs and documents in memory and does not require the startup PIN on wake. That is the case a lock screen bypass would attack.
- Limit physical access to unattended devices, and treat lost laptops as potentially exposed until there is a fix.
- Watch for Microsoft's response: an out-of-band update or the December 2026 Patch Tuesday, and the Microsoft Security Response Center page for any new advisory. Do not mix this up with the November BitLocker fixes (CVE-2026-97015 and CVE-2026-119832), which Cybersecurity News says are separate.
The researcher's suggested workaround
Per Cyber Kendra, quoting the researcher's X post, blocking Narrator with AppLocker or a custom access control list is a "potential fix" until a patch ships. Nobody I found has confirmed that this works, and Narrator is an accessibility tool: disabling it affects blind and low-vision users, so test in a pilot group and keep an exception path before applying it widely. I would not roll it out on one researcher's suggestion alone.
Why a lock screen bug is a security-boundary question
I found no statement from Microsoft on how it will classify a physical-access, lock-screen issue, so I will not guess. The practical point stands regardless: encryption at rest protects a powered-off device, and everything between "powered on" and "you typed your PIN or password" is a set of software boundaries that can have bugs. The earlier Hello Again attack on Windows Hello is another case in the same family. That is why the PIN-at-boot, shutdown-over-sleep and physical-security items above matter more than any single patch.
Context on the researcher
The reports note that this researcher has a history of publishing working exploits without coordinating with Microsoft, some of which were later exploited in the wild, and mention earlier work on the same theme: YellowKey (May 2026) and, per Cyber Kendra, a 2021 Narrator-driven BitLocker bypass for touch devices tracked as ZDI-21-1053. Security Affairs also reports that the researcher admitted an earlier claim about Microsoft legal action was fabricated. That is context for weighing claims, not evidence for or against this one. Judge the bug on whether it can be reproduced.
Frequently asked questions
Has Microsoft fixed it?
Not as of 20 November 2026 in the coverage I read. There was no CVE or advisory.
Is it being exploited in the wild?
No exploitation was reported in the coverage I reviewed. A public proof of concept exists, which raises the risk.
Does it need physical access?
Yes, according to every report.
Does TPM plus PIN protect me?
It raises the bar for the powered-off theft case and is the mitigation the reports recommend. It does not protect an unlocked or sleeping machine, and it is not a fix for Narrator. The researcher has not shown a bypass of it for this bug.
Should I disable Narrator?
Not on this evidence alone. It is an unconfirmed suggestion from the researcher and has accessibility costs.
update log
- 20 November 2026: first version.
- To recheck on each update: a Microsoft advisory, CVE or fix; independent reproduction (or failure to reproduce); whether it affects Windows 10 or TPM+PIN setups; whether Narrator blocking is confirmed to work.
Sources
- Tom's Hardware: New Windows 11 lock screen zero-day
- Security Affairs: Chaotic Eclipse Releases FastUnlock
- Cyber Kendra: FastUnlock Zero-Day Uses Windows Narrator
- Cybersecurity News: Windows 11 Narrator lock screen 0-day
- RuntimeWire: claimed Windows lock-screen bypass through Narrator
- ThreatLocker: YellowKey and GreenPlasma
Originally published on DEV Talk.
Top comments (0)