JupiteX Get the app
Science & Technology19 Nov 2026 · about 7 min

New Attack Bypasses TPM Lockout to Crack Windows Hello and BitLocker PINs in Days, Not Years

The brief

A TPM lockout is meant to slow or stop repeated wrong PIN guesses. Bypassing it lets an attacker reset or evade the recorded failure count, so the PIN can be tested far more often than intended. That matters because short PINs have relatively few possible combinations. In the researchers’ tests, they found a four-digit Windows Hello PIN and a six-digit BitLocker PIN after 236 tries each. The Hello laptop still restarted after every five incorrect entries, but the lockout delay did not apply. Each reset of the BitLocker test gave the researchers another 32 tries. These were lab demonstrations involving modified hardware, not remote attacks. The researchers say reproducing the attacks is not plug-and-play and requires one of their boards. They advise adding at least two digits to Hello and BitLocker PINs while AMD prepares firmware fixes for most affected chips between November and February 2027.

01

What does bypassing the TPM lockout let an attacker do to a Windows Hello or BitLocker PIN?

A TPM lockout is meant to slow or stop repeated wrong PIN guesses. Bypassing it lets an attacker reset or evade the recorded failure count, so the PIN can be tested far more often than intended. That matters because short PINs have relatively few possible combinations.

In the researchers’ tests, they found a four-digit Windows Hello PIN and a six-digit BitLocker PIN after 236 tries each. The Hello laptop still restarted after every five incorrect entries, but the lockout delay did not apply. Each reset of the BitLocker test gave the researchers another 32 tries.

These were lab demonstrations involving modified hardware, not remote attacks. The researchers say reproducing the attacks is not plug-and-play and requires one of their boards. They advise adding at least two digits to Hello and BitLocker PINs while AMD prepares firmware fixes for most affected chips between November and February 2027.

02

How much faster could the tested PINs be guessed with the lockout bypassed than with the lockout working?

The researchers calculated that bypassing the lockout could make a four-digit Windows Hello PIN about 41 times faster to guess in the worst case: 40 hours instead of 69 days. For a default six-digit BitLocker PIN, their estimate was about 347 times faster: 20 days instead of 19 years. These are comparisons of calculated worst-case times, not guarantees for every device.

In their demonstrations, the researchers set the PINs themselves. Their setup found the Hello PIN after 236 tries in 57 minutes, and the BitLocker PIN after 236 tries in 6 minutes and 55 seconds. The different times reflect the tested laptops and setup, not the worst-case estimates.

The speedup comes from defeating the TPM’s saved count of failed guesses, which should otherwise limit attempts. The researchers advise adding at least two digits to both kinds of PIN. AMD has confirmed a firmware flaw on some of its PCs and plans fixes for most affected chips between November and February 2027.

03

What physical access and equipment did the researchers need to carry out the attack?

The demonstrations required physical access and hardware work inside the laptops. The researchers did not simply try PINs through the normal sign-in screen: they removed the mainboard’s flash chip and connected their own board. Their methods interfered with the saved TPM state or with the flash chip’s counter protection.

A script read the screen, entered PINs, and restarted each laptop through a KVM switch. The researchers also changed firmware settings to slow the link between each laptop and its flash chip because their board could not keep up at full speed. For the Winbond method, they used a brief voltage glitch to extract a secret key.

The researchers have published their firmware, board designs, and scripts, but say reproducing the RPMC attacks is not plug-and-play and needs one of their boards. The article describes lab work; neither the paper nor vendor notices published so far describe attacks outside the lab. Opening the case alone does not describe all the equipment and steps used in the tests.

04

What could a recovered Windows Hello PIN unlock beyond signing in to the PC?

A Windows Hello PIN may protect more than the act of signing in to a PC. The researchers say a recovered PIN could also unlock keys tied to that device. This makes the issue relevant to accounts and services that use those keys, not only to access at the computer’s sign-in screen.

The researchers further claim that the PIN could enable Windows Hello-based single sign-on at work. In some configurations, they say it could also pass a multi-factor check. Those are claims about possible consequences; the researchers’ tests did not cover them. Their Hello demonstration recovered a PIN on a locked HP OmniBook 3, but did not test these other uses.

The article does not describe an attack on workplace accounts or a multi-factor system in practice. It says neither the research paper nor vendor notices published so far describe attacks outside the lab. AMD has confirmed a firmware flaw affecting some PCs with its chips and plans fixes for most affected chips between November and February 2027.

05

What is a firmware TPM, and how does it differ from a separate TPM chip?

A firmware TPM is a trusted platform module provided through processor firmware rather than a separate chip. In the PCs described, it has no secure storage of its own. It still needs to keep information such as the count of wrong PIN guesses, so that saved state is encrypted and stored on the mainboard’s flash chip.

A separate TPM chip differs because it has secure storage of its own, according to the research paper. The researchers’ attack targets the way firmware TPMs rely on saved state outside the processor. If someone copies that state and later restores the old copy, the recorded number of failed guesses can go back too, unless the PC detects the change.

Newer PCs try to detect restored data using counters inside the flash chip, called RPMC, that can only go up. The researchers found weaknesses in how some systems use these counters. Their findings vary by processor and flash chip, so the article does not say every firmware TPM or PC is vulnerable in the same way.

06

How can restoring old saved data or interfering with a counter make the TPM forget failed PIN guesses?

A firmware TPM saves its state, including failed PIN guesses, encrypted on the mainboard’s flash chip. If an attacker copies that data and writes an older copy back, the saved count can return to its earlier value. Unless the PC notices the rollback, it may accept more guesses as if fewer failures had happened.

Newer PCs use RPMC counters in flash chips to help detect old data; these counters are designed only to increase. On tested AMD PCs with RPMC flash, the firmware did not check that the counter actually increased. A small board could block the update while reporting success. The researchers also found that a voltage glitch could make some Winbond chips reveal a key used to sign counter values, allowing forged values.

The researchers found different weaknesses across systems. On the Intel laptop they tested, the firmware supplied the same counter value at each boot, so a leaked key kept working after restarts. Three Intel systems without RPMC-like flash detected write-back in the researchers’ tests. The results do not establish that every PC behaves alike.

07

Why does adding digits to a PIN make it much harder to guess by trying every possibility?

When an attacker tries every possible numeric PIN, each added digit creates more possibilities. If each position can be any of ten digits, adding one position multiplies the search space by ten; adding two multiplies it by one hundred. That makes trying every combination take longer, even if the attacker can make guesses quickly.

The article’s estimates show why this matters: with the lockout bypassed, the worst-case time was 40 hours for a four-digit Hello PIN and 20 days for a default six-digit BitLocker PIN. The researchers found each test PIN after 236 tries, but that does not mean every PIN would take that many attempts; the number depends on where it falls in the sequence.

Adding digits does not repair the TPM lockout weakness. It increases the number of guesses an attacker would need, while the intended lockout would also slow attempts. The researchers’ practical advice is to add at least two digits to Hello and BitLocker PINs. The article does not give a guaranteed attack time for longer PINs.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web