As artificial intelligence (AI) technology improves, it’s important for lawyers to understand how it connects with copyright law. In this blog post, we discuss four important subjects that lawyers should know about when handling copyright lawsuits and AI.
AI-Generated Works
A major development in AI and copyright is that AI applications can create new works, which raises questions about who owns the copyright. In the United States, the person who creates a work owns the copyright.
However, if AI generates a work, it’s not clear who the creator is. This can cause problems for copyright lawsuits and lawyers should know about these problems and how they could affect their clients.
AI-Assisted Infringement
AI can also be used to violate copyrights. For instance, AI can create Deepfakes which are edited videos or images that make it seem like someone said or did something they didn’t. These deepfakes can break the copyright of the real material and harm the reputation of the person in the video or image.
In the past few years, there have been many cases where people have been abused using deepfakes, an AI that is used to create fake pornography using non-sexual pictures and videos of real individuals. Actress Bella Thorne has spoken publicly about her own experience as a victim of deepfake abuse.
Helen Mort, a broadcaster, is another victim who was featured in stories by MIT Technology Review and the BBC and informed by the police that they couldn’t take any action because the current laws are old and don’t cover cases about copyright lawsuits and AI.
Fair Use and AI
The fair use doctrine lets people use copyrighted material without permission in certain situations, like for commenting, criticizing, or making parodies. But when AI is used to create these kinds of works, it raises questions about whether they still count as fair use. For instance, if an AI system makes a parody of a copyrighted work, who gets to decide if it’s fair use or not?
Data Training Sets and Licensing
Lawyers should also consider the licensing of data training sets used to create AI systems. These sets of data can be protected by copyright. Sometimes, the license for the data restricts its use to specific types of AI systems or applications. Lawyers should be aware of the terms in the license agreement for these data sets. They should also be mindful of the potential legal consequences if the data sets are acquired improperly, like through unauthorized access or theft.
Some companies, such as DeviantArt, Stability Ai, and Midjourney, have been sued by artists who claim that their artworks were used without permission to train text-to-image AI models. Stability Ai is also facing a lawsuit from Getty Images over similar allegations.
Conclusion
The intersection of AI and copyright law is a rapidly evolving area that lawyers need to be aware of. By staying up-to-date on developments in AI and copyright law, lawyers can provide their clients with the best possible representation and protect their intellectual property rights.
