Postulate one: while in the process of copying (or otherwise cloning) coded media, one can state that some minimal subset of the whole (not necessarily a workable subset of the whole) cannot constitute a meaningful proportion of the original media, meaning that it cannot be classed as a copy. Take the canonical case of a single bit of a movie file. At some point, the sliding window will represent some meaningful proportion of the original work, much like a collection of grains of sand becomes a heap, but there is established law in many societies that dictates fair use cases of subsets of media considerably larger than a single bit.
Postulate two: one can create a second work, larger than the original coded media, but containing the original coded media. If large enough, the original coded media can be contained in a form that is difficult or expensive to extract without relevant information, and potentially virtually invisible without relevant information.
Postulate three: if one exposes the entirety of the second work, yet restricts the relevant result of a copy or clone to only the sections of the media that correspond to the first work, then one can be said to have not copied the second work at all (by only allowing a restictive selection) but actually given access to the entirety of the first work.
Is junk DNA an IP circumvention mechanism?