I shall posit this for discussion:
There is sone excellent art work being included with some of the stories posted here. Yes, I know it's AI. Blah, blah, blah whatever. I suggest there should be a new category in the Clitorides Awards for art work.
Thoughts?
While you are at it, there should be a separate category for work written with AI.
How would that work? Do we rely on the SOL AI-generated tag, which may not encompass all AI-generated stories? Do other sites have a similar tag?
AJ
It would have to be marked as AI-generated, whether in the tags, or the description, or some other way.
I like the idea. It encourages authors to tag their AI-generated stories, so they're eligible for that award. Then the stories would be easier to identify.
I would suggest than anyone who uses AI to generate a whole story isn't an author.
Jo
I would suggest than anyone who uses AI to generate a whole story isn't an author.
Jo
Welcome to the party. This has been discussed and debated here for months.
I would suggest than anyone who uses AI to generate a whole story isn't an author.
I believe the work can't be copyrighted, so legally it doesn't have an author.
AJ
I believe the work can't be copyrighted, so legally it doesn't have an author.
That's not accurate. It has an author; there is just nobody who can enforce the right of publication.
I believe the work can't be copyrighted, so legally it doesn't have an author.
That's not accurate. It has an author; there is just nobody who can enforce the right of publication.
This may be overstating the state of US law a bit.
The US Copyright Office has issued an opinion that AI generated works can't be copyrighted, but AFIK, there are no US federal court cases directly on point.
My understanding is that the closest you can get is cases that decided that paintings by animals, and in one notorious case, photographs taken by a wild chimpanzee that got ahold of one of a nature photographer's cameras, can't be copyrighted.
The US Copyright Office has issued an opinion that AI generated works can't be copyrighted, but AFIK, there are no US federal court cases directly on point.
This may be overstating the state of US law a bit.
Not so. IANL, but the leading case is Thaler v. Perlmutter, the case definitively established the rule.
Computer scientist Stephen Thaler tried to register an image (A Recent Entrance to Paradise) that he said was created autonomously by his AI system, listing the AI itself as the author.
• Aug. 2023 — The U.S. District Court for D.C. (Judge Beryl Howell) ruled against him, holding that "human authorship is a bedrock requirement of copyright."
•March 2025 — The D.C. Circuit Court of Appeals affirmed, holding that the Copyright Act 'requires all eligible work to be authored in the first instance by a human being'. The court pointed to multiple provisions of the Act (inheritance by a 'widow', duration tied to 'the author's death', signature requirements, etc.) that only make sense if an author is human.
• May 2025 — The D.C. Circuit declined to rehear the case en banc.
• March 2, 2026 — The Supreme Court declined to hear the appeal, letting the ruling stand.
What hasn't yet been decided is how much human effort is required. From the case, I think we can draw the conclusion that a prompt is not enough. How much more is needed is an open question.
Worth pointing out that this only applies in the USA, and in other countries the position is different or unclear. The UK Government is currently consulting on this issue.
From the Consultation document:
Computer generated works
15. There are additional issues regarding the ownership of AI systems' outputs. The UK currently provides copyright protection for purely computer-generated works, but it is not clear that this protection is widely used, or that it functions properly within the broader copyright framework. The government seeks views on potential reform to protections for computer-generated works.
Worth pointing out that this only applies in the USA
Yes, of course, but given the original claim was about the US Copyright Office…
That's where I am on it, too, except I might say 'a simple prompt' is not enough. Whether complicated, detailed prompting is enough is an open legal question. Perhaps it also requires substantive editing after the LLM is involved; perhaps not. We won't know until it gets through the courts, and quite possibly not even then.
I suspect, in the meantime, we'll see the copyright office issue copyrights to humans copyrighting AI-written works without identifying it, which will only get adjudicated if there's a copyright lawsuit.
we'll see the copyright office issue copyrights to humans copyrighting AI-written works without identifying it,
I know this is pedantic, but worth pointing out that Copyright isn't 'issued' by any government agency, it's automatic upon creation of the work. Yes, you can 'register' your copyright for works created in the US with the Copyright Office (and other countries have a similar system), and there are significant advantages to doing so (it's essentially a prerequisite for litigation in the US), but it's by no means a requirement for Copyright to be valid.
In fact, the UK, Australia and New Zealand (that I'm aware of, there may be others) have neither an official Copyright office/agency nor any official way of registering copyright at all.
None of that negates the point that litigation and/or legislation in the area of AI and Copyright around the world will be worth keeping an eye on over the next decade or so.
Semi-agreed. The current thinking is that copyright is not automatic on creation of a work via LLM. The history above goes into what happened when someone tried to register a copyright.
If copyright is automatic when an LLM creates a work, then we're done, right? :) If you can't rely on a court finding there's a copyright, I would tend to say it's not automatic.
In which case, we're back to figuring out what a court (and/or the copyright office, for countries with those) will decide is actually protected by copyright.
The point I was making is that registration isn't required for Copyright to exist and that no government agency 'issues' Copyright. Indeed, in some countries around the world, it's not even possible to register Copyright.
The problem is the current arrangements were not designed with the current situation in mind. As I have said elsewhere, the British government has acknowledged that "wholly computer generated work" *is* currently protected by copyright as British law exists today, with the copyright belonging to the person operating the computer. There is no ambiguity in that statement, there is nothing for a court to decide. But the government has also acknowledged that the situation isn't fit for purpose as it stands.
What's required is legislation, and ideally international agreements, to clarify the position. Leaving it to the courts to decide piecemeal on a country by country basis, based on outdated laws, is far from ideal. What happens if we end up with LLM generated work being protected in the UK, but the not USA? Does the law apply from the country the prompt came from or the country the server resides in? How can we even know which country the server is in, given there are data centers all over the world? Is the situation then different if the LLM is locally hosted by the user or cloud based? Does it make a difference is the LLM model used was developed or is operated by a company based in the USA, China or the EU?
There are far too many variables and nuances for this area of (International) law to be left to a dispute between Bob in Alabama and Jack in Marine in a courtroom in Los Angeles. (I have no idea if that's how the US court system works, which is kind of part of the point I'm making).
we should add a motorcycle division to the Tour de France.
This is not a valid analogy. Motorcycles are not currently competing in the Tour de France. AI-generated stories are appearing on SOL and other story sites, and are already winning Clitorides awards. That's more like steroid-enhanced athletes in the Tour de France -- should those be in a separate division?
If motorcycles were already winning the Tour de France, I would approve of moving them to a separate category so there would be one motorcycle winner and one human-powered winner (and one winner in the steroid-enhanced division). This is not a problem because the exclusion of motors in Tours has been successful (so far as we know). If we discover that 80% of competitors are actually assisted by secret motors, it may become necessary to reconsider.
That's more like steroid-enhanced athletes in the Tour de France -- should those be in a separate division?
This. But also for every other sport. Don't forget that this year saw the first-ever "Enhanced Games".
This is not a valid analogy. Motorcycles are not currently competing in the Tour de France.
Not 'officially', but every time I accidentally see footage of the race, there are lots of motorbikes taking part.
Are there rules for 'drafting' behind camera and support (motor) bikes?
While we're at it, we should add a motorcycle division to the Tour de France.
Already have one, it just happens to take place on the Isle of Man.