Sorry but I will tell this former Mozilla engineer to silence himself right now. He worked on a web browser, not an AV product... He's got absolutely no idea. Does he even know how the injected DLLs work? They were not "rogue" in the way he is thinking of.
AV vendors don't follow standard security practices, which leads to many security bugs affecting the AV itself. To prove his point, O'Callahan points his readers to the
Google Project Zero project, and especially to the activity of Google security researcher Tavis Ormandy, who in the past two years has discovered gaping security holes in the software of many anti-virus vendors, which in many cases led to a complete takeover of the affected system.
You have to ask yourself: "Is this really the fault of the security product"? Sometimes it's not even in their control, but just due to gaps in the actual OS software. There are so many different ways to attack software it makes it impossible to prevent all attacks, and nothing is perfect... Obviously if someone is determined enough and has the skill-set then the target product is getting exploited no matter what you possibly do. This employee knew what he was doing due to experience and worked for Google, he wasn't a normal 20 year old on a blackhat forum copy-pasting code and compiling it, like most.
The person writing this article is making it seem like all security products can easily be tampered with, but this isn't the case what so ever; sure some security products might not be very good and may be vulnerable, but that doesn't mean they should speak for all of them. Kaspersky for example, have amazing self-defense mechanisms, and are not as easy to exploit... They use the damn hyper-visor which allows them to perform MSR hooking without worrying about PatchGuard protection for example, now tell me that is not advanced!
I bet this Google engineer can't even tell you how such technology works anyway.
AV products poison the software ecosystem because their invasive and poorly-implemented code makes it difficult for browser vendors and other developers to improve their own security. O'Callahan remembers that when Firefox implemented ASLR on Windows, AV vendors broke the feature by injecting rogue DLLs into the browser's process. Furthermore, several AV products blocked Firefox security updates for no apparent reason.
Firstly, does the writer even know the details on how the DLL was working? Secondly, does the writer know anything about signature/generic detection and white-listing?
Sure, the Anti-Virus vendor could have fully white-listed the vendor for Mozilla but maybe it was more sophisticated than this for the situation they were in, and if it was detection due to generic detection then it isn't a manual blocking. Then again, I wouldn't expect a browser software provider to understand how AV software really works, they can only dream.
They act like it's a crime for AV software to inject code into the other running programs, but what do they recommend as an alternate for the vendor to do so they can properly monitor the behavior of the running program and intercept/block before something bad really happens? I bet they can't recommend anything as a good and useful alternate, because they don't know what they're talking about...
It's hard for software vendors to speak out about these problems because they need cooperation from the AV vendors.
It's even harder when you are a browser developer and don't even know how to develop a security product.
I knew we were close, the day has finally come where I am being told to disable my security software from an engineer who dropped or was dropped by/from Mozilla, a browser developer, who lacks any real experience in the security industry.
What has this world come to....
Something is wrong here... I think he needs to see a doctor, having too much Adderal is bad for you and this is what happens when you do, and then start typing...