- Local time
- 7:42 PM
- Posts
- 39
- OS
- Windows 11 Home
Hi everyone.
It's me, the greenhorn/housewife/tech-dinosaur, trying to learn a bit more again.
Hoping for help regarding a driver bug. Here's the story:
I'm trying to install Win 11 Home 23H2 on a new computer we just assembled this month (first time DIY build, specs provided in my profile). @antspants kindly shared a couple of Win 11 Home ISOs with me, one in US English (build: 2861), and one in UK English (updated March 2025), so my husband could choose which one he wanted.
I used an up-to-date Google Chrome browser and the US English ISO downloaded without any interruptions. But some interruptions happened during the UK English ISO download, and I had to click 'resume' maybe 3-5 times to get it to complete. I've never downloaded large files like this before. I don't know if download interruptions can lead to any kind of corruption/missing stuff in the file or not. So just in case that might be a potential issue, we decided we would go with using the first (US English) ISO to make the bootable USB with Rufus. (The plan was to get it installed, do the 'tweaks' to make the system stay on 23H2 til EOL, then let the system bring 23H2 fully up to date).
So, in Rufus (the latest version, 4.7p) I did this:
Device: Kingston, 64 GB, NTFS.
(I did the sha-1 checksum thing mentioned on the Rufus github FAQ page and it showed a good match. Antspants is also sure the ISOs are fine, from a trusted source, and they have worked well for him. He still uses the US English one if he reinstalls Win 11.)
Device properties: Standard Windows Installation, GPT partition scheme, UEFI (non CSM) target system.
I clicked to enable runtime UEFI media validation. (A mistake maybe?)
Format options: file system-NTFS, cluster size-default.
I checked the advanced options for 'quick format', 'create extended label and icon files' and 'check device for bad blocks'. (There were no messages about any kind of bad blocks.)
Customization: I choose all except the region one and the one about secure boot & TPM since Rufus says secure boot doesn't have to be disabled anymore with its newest versions.
The USB booted the new computer easily and promptly. But then we got 3 warning messages on the screen. I'm not sure I fully understood what happened, but it seemed like maybe (?) it was because of me clicking to enable runtime UEFI media validation when doing the USB setup in Rufus? Because the new computer's screen warnings talked about md5 checksum stuff and referred to an AMI NTFS bug, telling us to read about it here: GitHub - pbatard/AmiNtfsBug: AMI UEFI NTFS driver bug test application
Seems like this driver bug creates vulnerabilities for persistent malware to get through the secure boot and settle in the UEFI, though it's only considered low risk of that happening. (Not sure what their definition of 'low risk' would be.) For us, though, we need to minimize risk as much as possible because we'll have to be doing sensitive work stuff on it as well as banking, government/tax, and healthcare stuff. So I feel a need to find out how to fix the vulnerability.
I cancelled the installation and then I tried to remake the installation USB with the 2nd ISO instead (the UK English one with march 2025 update). But I got the same notification from Rufus about a UEFI bootloader being revoked (which happened with the first ISO). That makes me think we'll get the same warning messages about the buggy AMI driver again, if we try the 2nd ISO for installation.
When I tried reading about the AMI buggy driver situation at the github link, as well as googling a bit and trying to search the forum here, I couldn't find hardly any info at all. What little there was felt like very confusing Greek to us and I couldn't find a simple layman's explanation about how to deal with the situation. We saw down near the bottom of that github page that it said, "using the UEFI NTFS driver from Releases · pbatard/ntfs-3g (unload the AMI NTFS driver and load the ntfs-3g one) does make the issue disappear." But I don't know how a person would unload or load these drivers in this situation. Not sure if the drivers are located in UEFI, or in Rufus, or in the Rufus-prepped USB. Don't suppose you have any tips for how to handle the buggy driver issue? or any links that might not have come up in my searches that could explain what to do, step by step?
In the release notes for the older Rufus 4.5 version, it said:
(Sorry for the long yarn about it. I wasn't sure how much info you would or wouldn't need, so I thought I should just tell it all...) Thanks for any help you might be able to provide. :)
It's me, the greenhorn/housewife/tech-dinosaur, trying to learn a bit more again.

I'm trying to install Win 11 Home 23H2 on a new computer we just assembled this month (first time DIY build, specs provided in my profile). @antspants kindly shared a couple of Win 11 Home ISOs with me, one in US English (build: 2861), and one in UK English (updated March 2025), so my husband could choose which one he wanted.
I used an up-to-date Google Chrome browser and the US English ISO downloaded without any interruptions. But some interruptions happened during the UK English ISO download, and I had to click 'resume' maybe 3-5 times to get it to complete. I've never downloaded large files like this before. I don't know if download interruptions can lead to any kind of corruption/missing stuff in the file or not. So just in case that might be a potential issue, we decided we would go with using the first (US English) ISO to make the bootable USB with Rufus. (The plan was to get it installed, do the 'tweaks' to make the system stay on 23H2 til EOL, then let the system bring 23H2 fully up to date).
So, in Rufus (the latest version, 4.7p) I did this:
Device: Kingston, 64 GB, NTFS.
(I did the sha-1 checksum thing mentioned on the Rufus github FAQ page and it showed a good match. Antspants is also sure the ISOs are fine, from a trusted source, and they have worked well for him. He still uses the US English one if he reinstalls Win 11.)
Device properties: Standard Windows Installation, GPT partition scheme, UEFI (non CSM) target system.
I clicked to enable runtime UEFI media validation. (A mistake maybe?)
Format options: file system-NTFS, cluster size-default.
I checked the advanced options for 'quick format', 'create extended label and icon files' and 'check device for bad blocks'. (There were no messages about any kind of bad blocks.)
Customization: I choose all except the region one and the one about secure boot & TPM since Rufus says secure boot doesn't have to be disabled anymore with its newest versions.
The USB booted the new computer easily and promptly. But then we got 3 warning messages on the screen. I'm not sure I fully understood what happened, but it seemed like maybe (?) it was because of me clicking to enable runtime UEFI media validation when doing the USB setup in Rufus? Because the new computer's screen warnings talked about md5 checksum stuff and referred to an AMI NTFS bug, telling us to read about it here: GitHub - pbatard/AmiNtfsBug: AMI UEFI NTFS driver bug test application
Seems like this driver bug creates vulnerabilities for persistent malware to get through the secure boot and settle in the UEFI, though it's only considered low risk of that happening. (Not sure what their definition of 'low risk' would be.) For us, though, we need to minimize risk as much as possible because we'll have to be doing sensitive work stuff on it as well as banking, government/tax, and healthcare stuff. So I feel a need to find out how to fix the vulnerability.
I cancelled the installation and then I tried to remake the installation USB with the 2nd ISO instead (the UK English one with march 2025 update). But I got the same notification from Rufus about a UEFI bootloader being revoked (which happened with the first ISO). That makes me think we'll get the same warning messages about the buggy AMI driver again, if we try the 2nd ISO for installation.
When I tried reading about the AMI buggy driver situation at the github link, as well as googling a bit and trying to search the forum here, I couldn't find hardly any info at all. What little there was felt like very confusing Greek to us and I couldn't find a simple layman's explanation about how to deal with the situation. We saw down near the bottom of that github page that it said, "using the UEFI NTFS driver from Releases · pbatard/ntfs-3g (unload the AMI NTFS driver and load the ntfs-3g one) does make the issue disappear." But I don't know how a person would unload or load these drivers in this situation. Not sure if the drivers are located in UEFI, or in Rufus, or in the Rufus-prepped USB. Don't suppose you have any tips for how to handle the buggy driver issue? or any links that might not have come up in my searches that could explain what to do, step by step?
In the release notes for the older Rufus 4.5 version, it said:
- Update UEFI:NTFS to latest (now always uses the ntfs-3g driver, rather than the buggy AMI NTFS one)
- Increase buffer size when copying ISO files, in an attempt to minimize the AMI NTFS UEFI driver bug
(Sorry for the long yarn about it. I wasn't sure how much info you would or wouldn't need, so I thought I should just tell it all...) Thanks for any help you might be able to provide. :)
Last edited:
My Computer
System One
-
- OS
- Windows 11 Home
- Computer type
- PC/Desktop
- CPU
- Intel Core i5-12600K
- Motherboard
- ASRock B760M PG Riptide
- Memory
- Crucial Classic DDR5-4800 16GB
- Monitor(s) Displays
- 1 good old Benq model
- Hard Drives
- Kingston KC3000 SSD 512GB PCIe 4.0 M.2 2280 NVMe
- PSU
- Seasonic G12 GM 750Watt
- Case
- metal, 15+ years old
- Cooling
- Thermalright Peerless Assassin 120
- Keyboard
- Lenovo
- Mouse
- Logitech, wired
- Internet Speed
- old DNS
- Browser
- Chrome
- Other Info
- First time DIY build.