Generative AI is helping businesses create content faster than ever before, but it’s also creating new copyright and intellectual property risks that many businesses don’t fully understand.
adobe stock
Generative AI allows businesses to easily create images, videos, marketing materials, and even their own bespoke software and tools. But there’s a big catch.
Much of the data used to train AI is created by humans. So if you come up with something that’s a little too close to someone’s trademark or copyrighted work, or something that violates a patent, you could get into legal trouble.
This has led to creative people suing AI tool developers for training AI on their work without permission and allowing others to create derivative works (e.g. Disney v. Midjourney).
But the most notable recent development is Alcon Entertainment’s pursuit of Tesla over its use of genAI imagery that closely resembles Tesla. blade runner 2049 IP suggests that end users increasingly find themselves in the crosshairs.
This is important because the amount of AI content being created is increasing every day, and is often made available to the public with little consideration for potential copyright issues. This suggests that many organizations still do not fully understand the risks.
Once again, we are witnessing a situation developing where technology is advancing too fast for the law to keep up. So let’s explore why this is dangerous and what you can do to protect yourself and your business from being taken to court.
AI plagiarism
The important thing to remember is that copyright infringement does not have to be intentional.
Many companies using genAI are probably not trying to copy anyone’s work. They are just looking for efficiency in content generation.
The problem is that the AI output may contain recognizable creative elements of other people’s work that appear in the training data.
Please note that there are two different issues with genAI copyright infringement. First, it is suspected that creative works were used illegally to train AI models. This is what a number of artists, writers, and musicians are currently suing AI companies for, as well as companies like Disney.
This is only a business issue if you are training your own model. The real risk for most companies is that the output of AI models may violate intellectual property regulations. No one knows what a court will decide on these issues in the future, and it is possible that a court will find, even by chance, that a copyright or trademark has been infringed.
As the use of AI evolves beyond content creation and becomes widely used across science, engineering, and design, the need to mitigate hidden IP infringement becomes even more important.
This makes Tesla’s lawsuit important because it directly targets end users.
Exactly how this will play out is unclear and will depend on further interpretation and clarification of intellectual property law by courts and legislators. This means that this situation is an ongoing business risk for those who use AI to generate content. So what are they doing to mitigate these risks?
What does this mean to you?
The golden rule is to always remember that AI itself cannot be held legally responsible for intellectual property theft. So are the companies and individuals who use it.
So there are a few things to keep in mind.
First, if you’re creating content that will be published, avoid giving prompts that encourage the AI to copy other content. Something like, “Please draw it in the Ghibli style.” This should go without saying, but it clearly isn’t.
Images, videos, marketing copy, and anything else made available to the public should be thoroughly reviewed before publication, with special emphasis on identifying elements that may resemble existing creative works or trademarks.
It’s also important to understand the coverage offered by certain AI service providers, such as Adobe, Microsoft, and Google. These provide access to models trained solely on your own IP or IP for which you have negotiated rights. This adds an extra layer of protection compared to simply relying on public tools where the origin of the data is unknown.
It is important to document how the AI content you publish is created to demonstrate that you respect intellectual property and have taken reasonable steps to avoid infringement.
Finally, keep in mind that the legal framework around AI and IP is evolving, and future court decisions and laws could significantly change what it means to use AI to create content.
However, if you are in a business where the speed and scale of generative AI creates tangible opportunities, you may not have the privilege of waiting until the situation is fully clear.
So while you don’t need to become a copyright expert overnight, you do need to learn how to work diligently and responsibly with AI content, and create processes to ensure it’s used safely and responsibly.
All of this should be covered in as much detail as possible in a company’s AI policy, minimizing the chance of mistakes slipping through.
But getting it right is key to using AI to create content at scale that doesn’t violate the rights of others or cause problems down the road.
