Under GPLv2 specifically the question is whether your business logic forms a derived work of the GPL-licensed software, which is nontrivial to interpret. My ro

Under GPLv2 specifically the question is whether your business logic forms a derived work of the GPL-licensed software, which is nontrivial to interpret. My rough understanding is that aggregating is permitted providing that the aggregation is merely that. For example, an image that can easily be extracted, for which it can be seen that busybox exists as a seperate binary, and for which the source code is made available by you (say, by virtue of the fact that you identify it as the unmodified BusyBox v1.31.1 and the source is on git.busybox.net), would be fine. Where it gets hazy is if you have some sort of proprietary image format going on and it cannot be verified that unmodified BusyBox is being distributed, or that you're not linking against it or using it as a substantial part of your logic. Bear in mind that the intent of the GPL, quite apart from how a lawyer would intepret it, is precisely that if you use GPL software in your software, you should release your software under GPL. If you're looking for loopholes then you're not really playing the game and should be looking for alternatives. My intent when writing GPL software is that it should not be included in proprietary solutions at all. I release my own personal projects under AGPL precisely because it's stricter. I don't ask for you to pay for my code in money - I ask for you to pay for it by providing the source code so that the wider ecosystem can benefit from your and my work.