CrowdStrike’s Falcon software uses a special driver that allows it to run at a lower level than most apps so it can detect threats across a Windows system. Microsoft tried to restrict third parties from accessing the kernel in Windows Vista in 2006 but was met with pushback from cybersecurity vendors and EU regulators. However, Apple was able to lock down its macOS operating system in 2020 so that developers could no longer get access to the kernel.
Now, it looks like Microsoft wants to reopen the conversations around restricting kernel-level access inside Windows.
In b4 msft creates a level between kernel and user level for this stuff to sit at. It will have read-only access to all of kernel memory, and will otherwise function the same, but when it crashes it won’t take the OS down, just certain programs that rely on it.
What will they call it? “Observer” level? “Big Brother” level? “Overseer” level? Probably just something to do with “Verifying Trust/Integrity”. Google will also want to quietly stick something for “Web Integrity” there.
WSS: Windows Subsystem for Stalkers.
https://github.com/microsoft/ebpf-for-windows
Right now it’s network level, but Linux’s implementation has since moved out from just packet filtering to full syscall filter and interaction; it’s generally accepted that Windows will be following suit with this implementation. Thought you’d like a name to the thing you described
Is it just me, or does this seem like a reasonable solution? Assuming it’s technically feasible.
It’s still giving third party software kernel level control over your device, so you’re still giving up any possibility of privacy and probably leaving yourself wide open to a backdoor attack, but that has been normalized. So to the degree that what people accept as reasonable these days is unreasonable, yeah, that’s why I think MSFT will try it.
This is exactly how ebpf was implemented for the Linux kernel. You can build watchdog processes that can see what’s happening in the kernel and build kernel interrupts but it’s actually all executed in user space and not rewriting the kernel itself. Since it’s a proper api, it also means it’s incredibly hard to fundamentally break the system, unlike when you’re just blowing away kernel code with your own shit like all these security products do.
That last line is some monkey’s paw type shit