27
ma𝕏pool (@maxpool@mathstodon.xyz)
mathstodon.xyzDoes copyright protect your AI-generated content in EU? Apparently not. Content that is entirely generated by artificial intelligence is not protected by copyright. EU copyright law has strictly human-centric foundation.
Daniel J. Gervais: 'When you put your name on an article that's written by ChatGPT or Claude, you're basically putting a provenance mark on it saying: I take responsibility for this. I haven't written it, but I'm putting my name on it. That doesn't give you copyright, but it does give you liability for the content' https://euobserver.com/232898/interview-does-copyright-protect-your-ai-generated-content-in-europe-lets-find-out/
Gervais, Daniel J. and Shemtov, Noam and Marmanis, Haralambos and Zaller Rowland, Catherine, The Heart of the Matter: Copyright, AI Training, and LLMs (September 21, 2024). Available at SSRN: https://ssrn.com/abstract=4963711 or http://dx.doi.org/10.2139/ssrn.4963711
"Munich Local Court has held that AI generated logos do not enjoy copyright protection. Neither mere prompting nor the selection between several AI suggestions is sufficient as a human creative contribution. For businesses, this is ambivalent. On the one hand, content generated purely by AI can hardly be protected on an exclusive basis, which has implications for brand building and content strategies" https://www.germanlawinternational.com/intellectualproperty/copyright/from-the-printing-press-to-ai-how-the-eu-plans-to-modernize-copyright-law-164154/
#law #copyright #LLM #AI #iplaw #intellectualProperty #EU
So if my AI model is dumb as fuck and just reproduces whatever comes to its input, but technically still produces AI-generated content… does it make it an uncopywriting machine?
IP attorney here that specializes in this stuff.
No, because that would be a reproduction of the original or a derivative work under copyright law, and those rights are retained by the copyright owner. Thus, it would be copyright infringement. The issue becomes fuzzier when you consider the temperature or entropy of a particular algorithm.
There is a lot more to this topic, but that is the gist that answers your specific question under US law right now.
I’ve seen that discussed as a genuinely serious problem for open source code. LLMs can be used to try to strip off a license.
That’s not how it works - the original author of the input owns the copyright for the work and if the output of the LLM is the same or a close derivative, the that copyright will be infringed. The ruling only means that YOU can not claim copyright for something that LLM spit out and you put your name on.
The key word in all this is “transformative.”
If you turn a thousand books into one book that’s not quite like any of the others… congratulations, that’s how writing works. Most authors read.
If you turn one book into the same book, that’s plainly not the same thing.