I think "could" is different from "will" - dna's existence is so heavily demonstrated that the chance of somehow disproving is is basically nil. But, the fact that we could disprove it (we look at it and don't see it there) helps us confirm its existence - in general in science something has to be falsifiable to be able to be demonstrated to be likely to be true if that makes sense, otherwise there's no way to obtain more evidence.

So it could be disproven, but will it? No. It definitely exists. But could it? Yep, that's how we know it exists in a scientific context. It's a bit counterintuitive, but it's the language of science. You see it in how physicists talk about discoveries - how many sigma they are. That's basically saying how likely it is the null hypothesis (sort of the falsification of the discovery) is true - and there's a point where it's so low (5 sigma traditionally) that it is considered "discovered." But that discovery is qualified with the chance that it's falsifiable.

I don't know if that was a clear explanation - if not let me know so I can try clarifying. It was something that took me a while to wrap my head around - that it's only demonstrable if falsifiable, but it's a core part of a lot of scientific theory today