Hey! Good news everyone! The #SupremeCourt, by declining to overrule the #Copyright Office, has declared that #AI art cannot be copyrighted.

It's bigger than just Thaler's case though. (links below) What this means is that NO AI created work can be copyrighted. Not code, not verbiage, not art, not anything.

Only human created works are eligible for copyright, so sayeth the High Court.

Cory (@pluralistic) does a fantastic job of explaining the details and rounding up links. As he always does. Where he finds the hours in the day, I will never know. ;)

https://pluralistic.net/2026/03/03/its-a-trap-2/

Archive link to #ArtNet because something on the page was blowing up my browser, but they have more details on the case itself, and the "art" in question which just saved real artists: https://archive.md/bsyrQ

Pluralistic: Supreme Court saves artists from AI (03 Mar 2026) – Pluralistic: Daily links from Cory Doctorow

@MissConstrue @pluralistic Im looking forward to the first people taking the windows OS code public

@MissConstrue @pluralistic AI generated code is public domain by default in USA legal jurisdiction?

That's going to get awkward. I guess trade secrets would still theoretically apply even if copyright doesn't.

It's still going to cause some interesting litigation.

@gooba42 @pluralistic

Right? RIGHT!? I need to stock up on popcorn.

@MissConstrue @pluralistic In which country?
@anne_twain @pluralistic I think it only applies to US? I can't imagine our courts have any jurisdiction anywhere else. Sorry for not adding a US tag in there.
@MissConstrue @pluralistic
This makes me VERY happy! 😃
@MissConstrue @pluralistic
It is not quite that simple. Anything that is entirely A.I. created cannot be copyrighted. As long as there is just a little human intervention/direct input, the work CAN be copyrighted. I read the Copyright Office document.
https://www.copyright.gov/ai/
Read part 2
Copyright and Artificial Intelligence | U.S. Copyright Office

Copyright law

@BoloMKXXVIII @pluralistic

Ah, but *only* the human created part can be copyrighted. That's the really fascinating bit about this.

So, for example, say I prompted an AI to give me a paragraph on current events. I then edit that paragraph, in the way that venture capital newspapers do, which is to say, lightly. Say I remove the em dashes, and delete some stuff it made up, and then I wrote the lede and the close, the only stuff that could be copyrighted is the lead, the close, and the edits.

As attorneys have explained it to me, the rest of the content remains without copyright protection. That said; I am not an attorney and I don't play one on the internet. Reading law is a specialized talent, and I may have misunderstood.

But, the Office says: the Office concludes that, given current generally available technology, prompts alone
do not provide sufficient human control to make users of an AI system the authors of the output...therefore no copyright on output.

@MissConstrue @pluralistic
I am not a lawyer either, but I did work with legal documents for almost 30 years. I seem to have a different interpretation than you do.

@BoloMKXXVIII @pluralistic

I figure the legal analysis by attorneys in IP and Copyright will have opinions soon. They're just more deliberate in consideration and celebration than I. ;)

It's going to be very interesting to see how it all shakes out.

@MissConstrue @pluralistic Wow - this is a big deal. For once it looks like SCOTUS gets it right

@MissConstrue
I wonder if that goes for code as well? We know the big tech AI just slurped up billions of lines of code from github gitlab codeberg etc.

@pluralistic

@justin @pluralistic There’s some question about that. My interpretation, having read the Office’s documents, is that code would be included in the “if not created by human, then no copyright” , but others reading the same document have reached different conclusions. I think we’ll have to wait for an IP attorney to weigh in for a solid answer.
@MissConstrue @justin @pluralistic But AI just reproduce what was written by humans. It's crazy that a copyrighted human code could become public domain when copy pasted by an AI 🤔

@wilk @justin @pluralistic

Then we start getting into "who stole the copyrighted work" and "How much do they owe for that".

I never gave permission for any of my work to be used to train AI, and yet...it has been.

So, if a github repository is slurped by openai, and then a junior coder asks openai a question that spits out content from that github, who should own the copyright? I would argue that the original creator owns the copyright, and therefore any subsequent use without permission is ineligible for copyright protection, as it is already protected via creation.

@MissConstrue @pluralistic looking forward to a case where an LLM reproduces verbatim some copyrighted text, for a company which has no knowledge of the prior copyrighted text, which then publishes it because they know LLM output cannot be copyrighted
@MissConstrue Yeah, but you can bet the corps will still find a way to make THEIR output copyrightable, or their terms of service to stand in for copyright. Ordinary people will still get the shitty end of the stick.
@cybervegan Yeah, the enshittification will continue until we rid ourselves of the enshitifiers.

@MissConstrue

Wow, historic ruling.

Does this mean if Microsoft Windows code is 30% AI generated, these parts need to be in the public domain?

https://pluralistic.net/2026/03/03/its-a-trap-2/#inheres-at-the-moment-of-fixation

@pluralistic

Pluralistic: Supreme Court saves artists from AI (03 Mar 2026) – Pluralistic: Daily links from Cory Doctorow