A recent discussion with a colleague about AI and copyrights, plus her subsequent article being published, led me to think about ChatGPT and copyrights.

The colleague is Melissa Rogozinski, who is the Founder of RPC Strategies LLC, a law firm and legal tech marketing company. Melissa is also a longtime legal professional in our industry and a longtime colleague. Many of us who know her call her “Rogo”.

Last week, Rogo published an article on her blog (How AI is Rewriting the Future of Marketing and Sales: Using LLMs Properly, Ethically, and Legally, available here), where she discusses some of her own experiences with Large Language Models (LLMs), which ChatGPT, Bard, and Claude are. She discusses some of the things she’s learned working with the technology in terms of what worked well (great to brainstorm ideas for all kinds of marketing and sales content) and not so well (ChatGPT-3 and Bard each provided an SEO strategy for a spa based on an old tag line – Rise | Renew | Regenerate, from 2019, instead of one better suited to her business).

Rogo provides several ways marketing and sales can leverage the power of LLMs (which are terrific and I won’t steal her thunder here), while reminding you to tailor the results for accuracy, quality, and voice before publishing in final form.

Another section of her article discusses ChatGPT and copyrights, which was a topic she and I discussed as part of her research for the article (in which, she references a blog post I wrote a couple of years ago about the Thaler v. Hirshfeld case, where the USPTO and the U.S. District Court for the Eastern District of Virginia determined that AI can’t be considered an inventor because the Patent Act defines an inventor as a natural person).

Can AI claim copyright over content it generated? Again, no, based on U.S. copyright law that Rogo references in her article, which was reinforced by Thaler’s recent attempt to also apply for a copyright on behalf of his DABUS system. Only works with human authors can receive copyrights, said U.S. District Judge Beryl Howell. And a human can’t claim copyright over something that is generative AI created, whether it’s a written publication, artwork, or anything else.

Makes sense. But what if content is influenced by generative AI, but not totally created by it? Is there a tipping point of how much AI is used before copyright can no longer be claimed?

For example, I can imagine an artist using Midjourney or DALL-E for ideas, then creating a picture themselves that is inspired by, or has similarities to, the AI-generated picture. Can that be copyrighted?

The same goes for written content. Let’s say you’re writing an article about the importance of metadata for eDiscovery, and you decide to ask GPT-4 a question about that, and it identifies two terrific reasons that metadata is important that you didn’t think of, so you decide to include them in your article along with reasons you already identified on your own. Can you still copyright the article?

What if you ask GPT-4 a question and it gives you such a great answer, you use much of what it provides to you as the bulk of the article, but with tweaks and rewordings to make it flow better. Can you still copyright the article then? Where do you draw the line with ChatGPT and copyrights?

Check out Rogo’s article here for great tips for using LLMs in marketing and also an interesting discussion on ChatGPT and copyrights (including the results of a poll on the topic).

So, what do you think? Does the inclusion of any AI content void the ability to copyright the work? Please share any comments you might have or if you’d like to know more about a particular topic.

Image created using Microsoft Bing’s Image Creator Powered by DALL-E, using the term “robot applying for a copyright using impressionism”.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.