The natural reading of a strange list is that it’s a set of examples and the courts will extend it sensibly. That reading is wrong. As one IP firm summarising the doctrine puts it, the list is short and comprised of a rather odd assortment of types of works capable of being treated as non-employee works for hire, and there are only nine categories. The consequence for commissioned art is blunt. A written agreement calling something a work made for hire is necessary but not sufficient – the work must independently fit one of the nine. A standalone photograph, a blog post, or a stand-alone musical composition commissioned from an independent contractor cannot qualify regardless of what the contract states.
That means many artist-brand agreements contain a clause that does nothing. The brand believes it owns the artwork outright. Absent a separate, properly drafted assignment, the artist may still hold the copyright.
Everything above applies to independent contractors. Where the artist is an actual employee producing work within the scope of employment, the analysis is entirely different and the employer owns the copyright from the moment of creation, with no list and no signature required.
Most disputes trace back to a contract that tried to do all three at once with a single sentence.
Here the money question gets interesting, because the statute contains a clawback most artists have never heard of. Under Section 203, an author who has granted away rights can terminate that grant roughly 35 years later and take the copyright back. The US Copyright Office describes the purpose plainly: the provisions exist to protect authors and their heirs against unremunerative agreements, giving them a chance to share in the later economic success of their work. Notice has to be served no less than two years and no more than ten years before the effective date.
Then comes the exclusion. The same Copyright Office page notes that grants involving a work made for hire may not be terminated under these provisions. So the comparison between two offers is not simply a comparison of two numbers. An artist who licenses or assigns rights for a smaller fee retains a reversion that arrives in year 35. An artist who signs a genuine work-for-hire agreement – one that actually qualifies – has no such right, because they were never the author in the first place and cannot terminate a copyright they never held. The larger cheque purchases the design and forecloses the clawback simultaneously.
For a graphic that sells modestly and is forgotten in three seasons, this is academic. For the small fraction of designs that become a brand’s recurring signature, licensed onward, reissued across a decade, or attached to something that turns out to matter culturally, it is the entire value of the relationship.
The duration numbers point the same way. A work made for hire is protected for 95 years from publication or 120 years from creation, whichever expires first. A work whose author is a human being runs for the author’s life plus 70 years. Depending on the artist’s age at creation, those two terms can differ by decades – and every one of those years belongs to whoever holds the copyright.
One detail catches brands out repeatedly. A work-for-hire agreement has to exist before the work is created. Courts have rejected agreements entered into afterwards that attempt to retroactively designate a finished work as made for hire, treating them as agreements to the contrary of the statutory termination rights. The practical version: a signature obtained at delivery, after the artwork exists, does not retroactively rewrite authorship. It is far too late by then, and the paperwork provides considerably less protection than the brand thinks it bought.
A licence can be limited by term, by product category, by territory, and by print run. It can carry a royalty on reissues. It leaves the artist with something to gain from the collection’s success, which tends to produce a more engaged collaborator and a cleaner relationship on the second project. And it avoids the awkward discovery, three years in, that the sweeping work-for-hire clause everyone signed was legally inert and ownership was never transferred at all. The buyout is simpler. It is also, for the work that turns out to matter, usually the more expensive option for whoever paid the higher fee – and the worse one for whoever accepted it.
