Prevent 32-bit scripts from having way more false positives than 64-bit

[lwc]

When using #AutoIt3Wrapper_Compile_both=y the resulting compiled scripts have huge differences in false positives.

First of all, Windows users' default antivirus (known as Windows Security's Virus & threat protection) actively quarantines 32-bit versions, while not caring at all about 64-bit versions.

Secondly, the different results in virustotal.com are very noticeable:

Script 1
32-bit - 10 false positives
64-bit - 1 false positives

Script 2
32-bit - 6 false positives
64-bit - 0(!) false positives

Script 3
32-bit - 7 false positives
64-bit -1 false positives

In fact, here are results of the most minimal script:

#AutoIt3Wrapper_Compile_both=n
msgbox("", "", "Hello World!")

32-bit - 12 false positives - see 32-bit scan results at shorturl.at/y9hL0 (linkable links aren't allowed here)
64-bit -1 false positives - see 64-bit scan results at shorturl.at/guIpz (linkable links aren't allowed here)

I tend to release apps as both 64-bit and 32-bit, since even nowadays it still seems common practice. But the 32-bit versions have been becoming quite a burden due to it.

Is there something in the compiler that can be improved for 32-bit versions?

[Jos]

This has been beaten to dead many times in the forum, so I honestly have no idea why you opened an BUG report being around since 2010!

[lwc]

Where in the forum was it discussed the differences between 32-bit and 64-bit where it comes to false positives?

[Jos]

Replying to lwc:

Where in the forum was it discussed the differences between 32-bit and 64-bit where it comes to false positives?

Don't know, but that shouldn't be a real surprise.

[lwc]

Replying to Jos:

Replying to lwc:

Where in the forum was it discussed the differences between 32-bit and 64-bit where it comes to false positives?

Don't know, but that shouldn't be a real surprise.

Why not?