Generative AI’s impact on ownership and copyright in AI-generated content is a subject of debate and legal challenges due to uncertainties surrounding human involvement and training data.
If you use artificial intelligence (AI) to create something, whose intellectual property is it: yours, the AI’s, or the owner of the content the AI was trained on?
The arrival of popular generative AI tools, trained on potentially copyrighted material from online sources generally without the consent of original creators, has added a new dimension to different forms of art and writing, sparking a conversation around this question.
The dangers of AI have been discussed thoroughly in the past few months and part of the ensuing debate has centred on the ownership and authorship of AI-generated content.
On Friday author and comedian Sarah Silverman, along with two other authors, started legal action against OpenAI and Meta for alleged copyright infringement.
Their lawsuit claims that the companies’ artificial intelligence models used their work in their training data without permission.
In addition to being a new tool in the tech world, AI has proven capable of completing tasks only humans could do until very recently. Now laws that have long protected people and their intellectual property are proving insufficient when AI is involved in the creation.
In the legal sphere, one of the main points of discussion are questions of copyright when it comes to images, texts and other forms of art generated by AI models via human input.
AI-generated art became a relevant controversy after an AI-generated work of art won the Colorado State Fair’s art competition in 2022. The piece was generated by Midjourney, a generative AI image tool, following prompts from artist Jason Allen.
The win triggered many angry responses from artists who claimed that AI will be the death of creativity and art if an AI-generated picture is considered more creative than human work.
While generating the right prompt for the piece demanded hundreds of different prompts from Allen – with the process as a whole taking more than 80 hours – the AI image was considered by many not worthy of competing with human creation.
Other than fear of AI taking over the art field and putting many creatives’ jobs at risk, an AI-generated image winning an art fair also raised questions around copyright issues. In other words, if all the artist did was come up with a description of the art but the AI tool generated it, who owns the rights to the generated image?
Who owns the copyright to AI-generated content?
In a recent paper published in Science magazine, researchers discussed copyright legal challenges when it comes to AI-generated work. They explained that there are lots of issues at play that make it challenging to copyright this type of content.
“The way that laws in the US especially work is that we have certain laws on the books and then when new technologies arise, then it’s up to judges to decide how those laws are going to be applied, and that means that we need to see cases. It’s such a new technology that a lot of the questions just haven’t reached the courts yet,” Robert Mahari, co-author of the study, told Euronews Next.
This is not the first time legal systems are faced with questions on how to define ownership of a specific creation where a new tool serves as a medium in its making.
Photography, for instance, was also an invention that altered our understanding of human creation back in the 1800s when it first emerged.
A court later ruled that photographers own the right to the photographs they create which led to photography becoming a new form of art on its own.
Perhaps this case in history is the closest that can serve as a guide on what to expect from AI-generated content copyright laws.
EU copyright rules for AI-generated content
Ever since the emergence of generative AI for public use, the EU has mainly been playing catch up in terms of regulations.
While the EU’s AI Act is the closest EU countries have gotten to regulate AI use, specific laws for copyrighting AI-generated content are yet to be established and until now it might just be dealt with on a case-by-case basis.
However, part of the AI Act that members of the European Parliament have recently voted on addresses the need for companies that deploy AI-generated content to disclose any copyrighted material that was used to develop their systems.
Considering one of the biggest challenges to copyrighting AI-generated content is the possibility of copyrighted material being used to train the AI system, labeling it could be a step in the right direction that will potentially lead to more refined copyright laws in relation to AI-generated content.
What are the challenges to copyrighting AI-generated content?
The main reasons why it’s challenging to copyright AI-generated content are the ambiguities regarding human involvement and intentions, the training data the AI tool could have used in generating the output, and various questions regarding ownership.
“Currently for art, this data is scraped from the web and it is scraped without asking artists for their consent, without really notifying artists that the data is going to be scraped,” Mahari said.
The data that generative AI is trained on includes many creative works that are protected under copyright laws, and that, in most cases, were included in the training of the AI system without the creators’ knowledge or consent.
“So you use the training data to train this model and then in the final step, you have a trained model and then you can use it to create new outputs. Now, even the first step, even just taking the data and training an AI model can raise copyright issues because you’re now transforming this art into something new,” he said.
In US copyright law, there exists the notion of ‘fair use’ which basically allows creative work based on a copyrightable artwork – but it should be transformative enough that it’s somewhat different from the original. This type of altered work is considered separate from the original work of art and is not subject to copyright infringement.
According to Mahari, there are still many questions regarding whether an AI tool generating a piece of art copying the style of an original work is to be considered fair use, and even in the case where it is, there will still be a strong need to compensate the original artists and protect their work.
While AI doesn’t always generate responses inspired or based on other artists’ work, it is hard to trace back the training data the tool used to generate the output.
Many lawsuits have already been filed against AI image generators that contain copyrighted images in their training data.
Beyond the complications to determine what copyrighted work went into generating a specific piece of AI content, another struggle would be determining how much human involvement exactly is enough for the work to be owned by them.
Since copyright laws mainly aim to encourage art and its creation by protecting artists and their unique creative ideas, lawmakers might need to consider joint ownerships.
“We could extend copyright law and we could consider things like joint ownership,” Mahari said.
“Even though copyright law currently doesn’t protect styles, if you create artwork that’s very similar to someone, maybe they kind of have a joint ownership,” he added.
Considering that, as of today, there are no copyright laws that address AI-generated content specifically, and the situation is still solved on a case by case basis in court, people may need to consider how they are using generative AI in the process behind their creative work to avoid copyright issues later.