Those claiming AI training on copyrighted works is “theft” misunderstand key aspects of copyright law and AI technology. Copyright protects specific expressions of ideas, not the ideas themselves. When AI systems ingest copyrighted works, they’re extracting general patterns and concepts - the “Bob Dylan-ness” or “Hemingway-ness” - not copying specific text or images.
This process is akin to how humans learn by reading widely and absorbing styles and techniques, rather than memorizing and reproducing exact passages. The AI discards the original text, keeping only abstract representations in “vector space”. When generating new content, the AI isn’t recreating copyrighted works, but producing new expressions inspired by the concepts it’s learned.
This is fundamentally different from copying a book or song. It’s more like the long-standing artistic tradition of being influenced by others’ work. The law has always recognized that ideas themselves can’t be owned - only particular expressions of them.
Moreover, there’s precedent for this kind of use being considered “transformative” and thus fair use. The Google Books project, which scanned millions of books to create a searchable index, was ruled legal despite protests from authors and publishers. AI training is arguably even more transformative.
While it’s understandable that creators feel uneasy about this new technology, labeling it “theft” is both legally and technically inaccurate. We may need new ways to support and compensate creators in the AI age, but that doesn’t make the current use of copyrighted works for AI training illegal or unethical.
For those interested, this argument is nicely laid out by Damien Riehl in FLOSS Weekly episode 744. https://twit.tv/shows/floss-weekly/episodes/744
I’ve never done this. But I have taken lessons from people for instruments, listened to bands I like, and then created and played songs that certainly are influences by all of that. I’ve also taken a lot of art classes, and studied other people’s painting styles and then created things from what I’ve learned, and said “look at what I made!” Which is far more akin to what AI is doing that what you are implying here.
So what if its closer? Its still not an accurate description, because thats not what AI does.
Because what they are describing is just straight up theft, while what I describes is so much closer to how one trains and ai. I’m afraid that what comes out of this ai hysteria is that copyright gets more strict and humans copying style even becomes illegal.
I’m sympathetic to the reflexive impulse to defend OpenAI out of a fear that this whole thing results in even worse copyright law.
I, too, think copyright law is already smothering the cultural conversation and we’re potentially only a couple of legislative acts away from having “property of Disney” emblazoned on our eyeballs.
But don’t fall into their trap of seeing everything through the lens of copyright!
We have other laws!
We can attack OpenAI on antitrust, likeness rights, libel, privacy, and labor laws.
Being critical of OpenAI doesn’t have to mean siding with the big IP bosses. Don’t accept that framing.
Well that all doesn’t matter much. If AI is used to cause harm, it should be regulated. If that frustrates you then go get the laws changed that allow shitty companies to ruin good ideas.