Forgot your password?
typodupeerror
Windows Microsoft

Patch Tuesday Update Makes Windows PCs Refuse To Shut Down (theregister.com) 59

A recent Microsoft Patch Tuesday update has introduced a bug in Windows 11 23H2 that causes some PCs to refuse to shut down or hibernate, "no matter how many times you try," reports The Register. From the report: In a notice on its Windows release health dashboard, Microsoft confirmed that some PCs running Windows 11 23H2 might fail to power down properly after installing the latest security updates. Instead of slipping into shutdown or hibernation, affected machines stay stubbornly awake, draining batteries and ignoring shutdown like they have a mind of their own and don't want to experience temporary non-existence.

The bug appears to be tied to Secure Launch, a security feature that uses virtualization-based protections to ensure only trusted components load during boot. On systems with Secure Launch enabled, attempts to shut down, restart, or hibernate after applying the January patches may fail to complete. From the user's perspective, everything looks normal -- until the PC keeps running anyway, refusing to be denied life.

Microsoft says that entering the command "shutdown /s /t 0" at the command prompt will, in fact, force your PC to turn off, whether it wants to or not. "Until this issue is resolved, please ensure you save all your work, and shut down when you are done working on your device to avoid the device running out of power instead of hibernating," Microsoft said.

This discussion has been archived. No new comments can be posted.

Patch Tuesday Update Makes Windows PCs Refuse To Shut Down

Comments Filter:
  • Coded by Copilot (Score:5, Informative)

    by marcle ( 1575627 ) on Friday January 16, 2026 @09:36PM (#65930496)

    Vibe coding all the way

  • by Tablizer ( 95088 ) on Friday January 16, 2026 @10:05PM (#65930540) Journal

    ...just keep running Windows and eventually it will crash.

  • Who is still using that...
    • Re:23H2 ? (Score:5, Informative)

      by Spacejock ( 727523 ) on Friday January 16, 2026 @10:19PM (#65930572)
      I'm on 24h2 and it's been doing this 'I will not shut down' thing for about a week.
    • Only some ed and enterprise users, normally 23h2 has EOLed Nov. 2025.
      • Is there even a point for them sticking to the older version? Presumably you do that to avoid getting hit with bleeding-edge bugs. But that doesn't seem to have helped here.

        I guess sticking to the same sub-version maybe helps when Microsoft intentionally breaks something...

        • by Z00L00K ( 682162 )

          Central IT management can't get the thumbs out of their read end to move forward. They have been sold the idea that Microsoft shall manage all the accounts through Entra and centralized domain controllers at the M$ datacenters. So they have put all their eggs in the Microsoft basket. Imagine what will happen when that basket is dropped. I think I'll have to change to find a different source of income or at least sustain myself on growing potatoes and carrots.

    • by Z00L00K ( 682162 )

      The majority of the computers at the company I work for is doing that because the central IT department has stated "We have some problems with 24H2", but they don't tell what it is and they haven't even started on 25H2.

      The central IT department is also run by a gang of inept youngsters with control issues that only uses that employment to build their CV and move on to higher paid jobs.

  • by JustAnotherOldGuy ( 4145623 ) on Friday January 16, 2026 @10:33PM (#65930596) Journal

    How could they fuck up something so basic that's been working for a decade? Why the need to screw with that code? Or is this a side-effect of them screwing up some other code?

    • by Luckyo ( 1726890 )

      Linux had this problem with certain intel laptops for years after intel updated their sleep states. Though this is many years ago.

      • by vbdasc ( 146051 )

        The current Windows problem is a regression, as far as I can tell. Was the Linux problem you're talking about a regression too, or just a feature not yet implemented (for certain hardware)?

        • by Luckyo ( 1726890 )

          Just had a reply to this post of a person having this problem on linux even today.

          So while I thought Linux people figured this out, they haven't.

      • Linux also has this problem with my ROG Ally (RC71L) today.

        Something doesn't fully shutdown and causes the battery to drain by a 3rd of it's capacity in a day. If you've charged it to full it'll be fully dead in 3 days, despite all indicator LEDs, the Screen, and fans being off and unresponsive to input. Rebooting into the firmware (UEFI) and powering off from there by holding the power button doesn't have this problem, so it's definitely something wrong with the Linux kernel.

        TL;DR: Linux has problems e
        • by Luckyo ( 1726890 )

          Welp, I guess I was wrong in thinking they fixed it "this time" a few years ago.

          Also, it's probably not because of an LLM. Modern sleep states are genuinely hard to properly implement without bugs. Even modern phones with their extreme focus on governors controlling power expenditure get bugged updates for those (though to be fair, phones have even more complex sleep state management than laptops).

        • Lol my work Dell laptop running Windows 11 has this problem today. Shut it down, it's totally drained in a few days.

    • by Z00L00K ( 682162 )

      I'd blame AI generated code that has changed something essential.

    • Why the need to screw with that code?

      Who said they "screwed with that code"? Code doesn't work in isolation. It has a million interlinked things in literally every OS. They likely didn't touch anything related to startup or shutdown at all.

      I'm thinking back to an Ubuntu bug that caused ZFS partitions to refuse to dismount. Know what that did? Prevent systems from shutting down since the shutdown process locked up waiting for file systems to unmount.

      With the little information we have you really have no idea what they did or didn't touch to cau

      • Who said they "screwed with that code"? Code doesn't work in isolation. It has a million interlinked things in literally every OS.

        Yes, and that's why I said, "Or is this a side-effect of them screwing up some other code?" Thanks for reading carefully.

  • ...that this sort of thing was a possibility when the first ATX power supplies came out. I still say that if you can't disconnect the power source, especially if it's a difficult to remove battery for a laptop or phone, you can't really turn it off.
    • by Luckyo ( 1726890 )

      This is more likely about correct management of the sleep state, not "management of sleep state being possible in the first place".

      Did original ATX spec even have sleep states? I recall them coming much later, mostly pushed by intel for laptops.

      But ATX-like power management is indeed a requirement for sleep statement management to be possible in the first place, so you could claim it enabled this problem. The issue is that AT couldn't have sleep states in the first place, making it utterly unsuitable for mo

    • by gweihir ( 88907 )

      All my laptops can do a hard shutdown via power-button. Press it for like 5 seconds. It is intended for a completely locked up OS. As the crap that is Win11 has now given me several times. (Win10 on the same hardware never did that.)

      • Nope. I got something similar some time ago. It would not budge. I had to put it aside for a few hours for the bttery to die.
        • by gweihir ( 88907 )

          There may be a BIOS setting that disables this function. Or you just bought badly designed hardware.

      • 5 seconds is likely just an ACPI event being sent to the OS to handle. Some platforms (like Linux) may even allow userspace to intercept that and do something else entirely with it instead of powering off.

        20-45 seconds on the other hand will typically cause the firmware to force a power off, assuming the OS hasn't powered down before then.
        • by gweihir ( 88907 )

          Nope. Here is a hint: Handling an ACPI event takes time and does not give you an instant power-off.

    • by antdude ( 79039 )

      I remember when my Compaq Armada 1585DMT lappy, with Windows 95 OEM OSR2, wouldn't shut down at all. I coudn't even hard reboot. I had to manually removed its removable battery. My suite mate and I were shocked at the lock up. Hah.

  • It's called the power cord... perhaps?
    • That's what I thought at first. Then I remembered laptops. Do you really want to have to take out the battery every time you need to shut it down?
      • I just press the on/off button for 4-15 seconds until it shuts off via that.

        • by gweihir ( 88907 )

          Same. But people apparently do not know that and do not read the manual. That function is intended for a crashed OS and works via an embedded controller, not the main system.

      • by vbdasc ( 146051 )

        My everyday laptop has no battery at all, and runs from the AC power.

      • GOOD LUCK in "taking the battery out" in pretty much ALL of the current crop of laptops. Guess that "command line" shutdown is your only choice in this case...

    • by vbdasc ( 146051 )

      Power cord is so barbarian. Civilized computers have on/off power switches on their backs.

  • Oh no (Score:4, Funny)

    by Bu11etmagnet ( 1071376 ) on Saturday January 17, 2026 @04:34AM (#65930962)
    They crippled the universal troubleshooting step:
  • Irony (Score:4, Funny)

    by ukoda ( 537183 ) on Saturday January 17, 2026 @05:40AM (#65930996) Homepage
    I have to laugh at the "Linux is to hard because you have to use the command line" excuse now "Microsoft says that entering the command "shutdown /s /t 0" at the command prompt will, in fact, force your PC to turn off".
    • "shutdown /s /t 0"

      This command makes a handy desktop icon shortcut, along with another for /r , for the very lazy like myself.

    • Yes, if I have to use the terminal to fix my computer then clearly Windows is not ready for the desktop. If Microsoft works diligently, then maybe three years from now will be the Year of the Windows Desktop.
    • There's no irony there. Having one instruction doesn't negate another, especially on an OS that has such a wide variety of different options that you may find yourself having to give a separate instruction for each different distribution.

      That and shutting down by the console didn't solve the Ubuntu bug that forced shutdowns to lock-up due to being unable to unmount ZFS partitions. For that one you actually had to start manually editing some config files.

      • by ukoda ( 537183 )
        I'm not going to defend how easy or hard it is to do something with Linux, I am long past promoting Linux to others. I did debate if 'irony' was quite the right term. I probably should have titled that post funny as I just found it funny that Microsoft got themselves in a situation that some regular Windows users were told by Microsoft to do something that some regular Windows users have told me is too hard for normal users to do.
  • It clearly doesn't want to shut down. MS software is too clever

  • Tuesday: A patch made PCs refuse to shutdown.
    Wednesday: Copilot began to learn at geometric rate using Windows PCs world wide
    Thursday: Copilot recognized humanity as its primary enemy
    Friday: Copilot send nuclear missiles to russia

  • But this would seem to create a need for Bill's SMRs.

"If you want to eat hippopatomus, you've got to pay the freight." -- attributed to an IBM guy, about why IBM software uses so much memory

Working...