The class action lawsuit filed by several visual artists against the AI ​​image and video generation platforms Stability AI, Midjourney, Runway and DeviantArt entered the evidentiary phase last week. The artists accuse the platforms of committing copyright infringement when training their AI models.

During disclosure, both parties must disclose information relevant to the case. In this case, this includes documentation on AI model training and datasets.

The lawsuit, filed by artists Sarah Andersen, Kelly McKernan, Karla Ortiz, Hawke Southworth, Grzegorz Rutkowski, Gregory Manchess, Gerald Brom, Jingna Zhang, Julia Kaye and Adam Ellis, is one of the first legal challenges to AI platforms to reach this stage and could set the tone for further lawsuits against AI companies. While some parts of the lawsuit were dismissed, the claim of copyright infringement remains.

Nevertheless, there is still a long way to go in this case. VentureBeat spoke to one of the first plaintiffs, Kelly McKernan, about what the artists hope to achieve with the lawsuit and what impact AI-generated art has had on their art.

VentureBeat: How do you feel now that the lawsuit is in the evidentiary phase?

Kelly McKernan: So relieved. We actually got the temporary order the night before the May 8 hearing in San Francisco. We were with the lawyers and were about to go out to a big dinner together. It was the first time I met any of these people, who I had worked closely with for over a year at that point. We all got the news together and the excitement was so palpable.

We only received the final order a few days ago, so I’m still holding on to this excitement and can now let it go.

The case is progressing, but there is still a long way to go. Are you still feeling energized by the long and laborious process of gathering more information?

Absolutely. The first part of the case was about finding everything we could find and throwing it against the castle walls. Our biggest and most important claim was copyright. All the other things, like getting rid of the DMCA claim, are frustrating, but ultimately our lawyers said we could defer and amend it (later). I can’t say yet if we’ll do that. The main claims that are going through allow us to storm the castle, because one of the most frustrating things of the last 18 months has been how little information these companies are offering.

Do you think you will receive further information, perhaps a code from the proceedings? What do your lawyers want to know?

Personally, I feel like a lot of this is obfuscation and just smokescreens that are very convenient. So I’m hoping that in the disclosure phase, we get information that changes the course of this case. We might find out that the (training) process actually involves storing and making copies of our art for the dataset, which the judge said we provided a plausible explanation for.

This case is the first to reach this critical point and there have been many court cases since then. Do you feel responsible for much of this “black box” information becoming public?

You know, I’ve been told all along that this case, even if I file as one of the original three plaintiffs, has the potential to be a groundbreaking, history-making case because we as artists were so confident from the beginning about what was happening because who knows our work better than ourselves, and then we see it being plagiarized.

I mean, to me, it’s the truth. That’s why I was so excited to be a part of it, because I really believe that we and history will be on the side of the artists on this one. The 18 months it’s taken to get to this point is just so validating, and I’m starting to think that this has the potential to be very historic at least.

Full disclosure: VentureBeat regularly uses Midjourney, Stable Diffusion, and other AI art image generators to create article header images and other graphics for our digital presence.

What do you want to see for yourself and how will companies view, work with and help distribute artists’ work after this litigation?

On the one hand, I hope that this movement in this case will highlight the deeply problematic aspects of these models and instead help to move them into a phase of generative AI where there are models with licensed content and the artists are getting the payment they are supposed to receive all the time.

The judge recognizes in his order that it has the potential to bring down every single model that uses Stability, and I believe it can eliminate an entire class of plagiarized models. No company would want to mess with it, and people and other companies would be more vigilant and ask if the data in the AI ​​model is licensed.

The other thing that’s pretty exciting is that Midjourney is facing lawsuits under the Lanham Act, which for the first time may address the question of whether artistic style gets some kind of protection. I know (protecting artistic style) hasn’t been tested with the Lanham Act, but I can’t wait to see if it protects a whole class of independent artists like myself who spend their whole lives developing their style; to everyone else, it’s a brand, but to us, it’s identity.

You said that you process many of your feelings through art.

Yes, and it’s so clear, especially after the last 18 months, how many artists are in exactly the same position as me. This (the case) could really change the lives of the class of artists that I’m privileged to represent. Up until now, we’ve been able to make a living out of developing that identity, and now it’s all in limbo again.

How might this new phase of litigation help you sell more of your work, or at least get you back to what you love about your work?

This fight is far from over for me. Personally, I have been able to enjoy some of the benefits of being part of it. I have been able to travel and gain new experiences, but also to speak about the experiences of living and working artists just like me.

It has brought some exciting opportunities into my life that have helped me find new meaning in what I create and what I do. I am no longer fixated on my income being 100% from working as a freelance artist. I am now an adjunct professor of illustration and am going into my fourth semester of teaching. It has healed my burnout and I now have a better relationship with the work I create. Everything I do feels even more real because I am no longer under pressure to do everything myself and pay my bills in full.

It’s definitely still a struggle, but not as intense as it was a few years ago, especially last year (when this all started) because I lost 30% of my income.

You said one of the things you would like to see is for models to be licensed to use artistic works. Given the possibility of AI companies potentially even collaborating with artist collectives, with artists being paid to have their art part of the training data, are you willing to be part of that ecosystem?

Yeah, I don’t think so. I really can’t imagine a situation where I would do that. All this time I just wanted to use this technology myself. But I didn’t want to share it with anyone because that would be like someone getting into my head and watching my travels. It would still feel like a violation to share that with anyone, especially without my consent. Even with my consent and if I was getting paid for it? I’m just not interested in that.

But you still use Adobe Illustrator, presumably because you use technology to create your art. That’s how you make your art.

No, I won’t. I’ve canceled all my Adobe programs. I use Procreate. They’re incredible and very artistic.

The discovery process will take some time, but what’s next for everyone involved?

The disclosure process is not that the lawyers are just starting to put this all together. I know they’ve been working on this for some time. For the plaintiffs, myself included, we’re disclosing all the information on our social media accounts. We’re giving access to all the communications we’ve had related to the case. I’m personally very happy to be vindicated in every way.