As artificial intelligence continues to create increasingly sophisticated content, from music and art to news articles and videos, questions about copyright ownership and infringement are becoming more urgent. With machines able to generate original works, who owns the copyright and how can it be protected? These are just some of the challenges facing lawmakers, creators and technology companies as they grapple with the legal and ethical implications of AI-generated content.
And all the above was written, by the way, by Open AI’s ChatGPT technology.
Businesses large and small are able to generate content — derived from basic prompts such as was used for this article’s introduction — with no more labor than a few keystrokes and clicks. While the process might be easy, Wendi Uzar, an intellectual property partner at Riker Danzig in Morristown, believes those business owners should be also aware of the complex legal issues surrounding the use of this content.
While it might sound esoteric to some, Uzar said it’s something that’s becoming a popular topic of discussion for intellectual property lawyers. More each day, she’s asked questions about the legal realities behind AI’s impact.
Here’s how she’s answering those queries. …
ROI-NJ: First off, how does current intellectual property law square with the rise of AI-generated content?
Wendi Uzar: It’s a complicated question right now. As of Feb. 21, the (U.S. Copyright Office) issued an opinion letter in relation to a case involving images that were created by AI. And the agency in the letter stated that content created fully by AI cannot be protected by copyright. However, if content was manipulated by a human hand to an extent that it meets the standard of sufficient creativity, then it could possibly obtain that protection. But that’s all still unfolding. What’s enough manipulation to maintain that standard? How involved in the selection and coordination of that work do you have to be to meet that standard? Will you eventually be able to protect AI-generated content? Right now, there’s not even a requirement to say you used (an AI technology). All these issues are still being fleshed out.
ROI: Who would actually be at fault for copyright infringement if content generated through one of these seemingly hands-off platforms itself was the issue?
WU: If a business uses AI to create content found to infringe copyright, then that company would be liable. So, similar to when you use any third-party to source images, it’s important to do homework before you choose what company you want to work with. And you want to check your agreement carefully. You want to look for indemnification clauses to protect your business. If you use a company and are later charged with copyright infringement, then hopefully you can go back to that company and use that clause to hold you harmless in that case. But it’s also important that the company is credible. And I also advise clients to ensure those companies are based in the United States. Because a company may say they’ll indemnify you, but it’s hard to enforce that if a company is located abroad, in my experience.
ROI: Is it a problem that AI-generated content sometimes seems to resemble existing content that might already be protected by copyright?
WU: I certainly think that’s an issue. Because, on the other side of the coin, what if an artist finds out their material is being used through AI as a basis to create new images or data by AI? For example, if someone uses a program and says, ‘I want a painting similar to this specific artist.’ If that’s the starting point, then that could in and of itself be considered derivative work and copyright infringement. I always advise clients whenever putting anything online to use a watermark over an image so it’s not so easily grabbed and potentially used again in future.
ROI: Given all that, should business owners even fuss with this?
WU: My opinion on this is to be cautious about putting out content using (AI technology). With the rules saying you cannot obtain protection on it right now, you don’t want to utilize time and money to generate content using AI, put it out on your website and have third parties freely copy and utilize it. Unfortunately, copyright law doesn’t take into account how much money you spent creating content like that. In general, my advice right now would be to wait on this. It’s so new that there’s still determination being made on the best path forward on how to manipulate this data and how to protect it. If you’re going to use it, I would strongly advise to manipulate it significantly — and document that process, too. It should probably only be a springboard, not something you’re putting out directly as it’s spit out from AI. That way, you at least have a stronger argument for protecting it and stopping others using it.
ROI: Any other suggestions for business owners not yet scared off of an embrace of these new technologies?
WU: Businesses should also have a policy set forth on what the expectation is if you’re allowing employees and independent contractors to use these AI tools. You want to make sure you’re advising employees on where you stand on it. If you do allow the use of AI content, it’s advisable to have one account for the company and someone responsible for going in and monitoring use and ensuring employees follow guidelines on that account.