The reception to ChatGPT has been so enthusiastic among some users that intellectual property lawyers are finding that they must walk a fine line between warning clients about possible legal issues with the groundbreaking artificial intelligence tool and respecting their ardor for the product.
Since its rollout in late 2022, ChatGPT has been riding a tidal wave of media coverage, with reports that it passed the bar exam and could replace humans in some professions.
Fans of the platform speak of its ease of use, versatility and language skills, and its ability to design a logo, draft a technical manual or write a poem.
But businesses embracing ChatGPT should know about the complex issues around ownership of content created with AI, according to some attorneys.
Human Touch
The U.S. Copyright Office issued an advisory on March 16 stating that only content produced by humans is eligible for copyright protection, and works that are the product of AI are not.
But the office said in some cases, “a work containing AI-generated material will also contain sufficient human authorship to support a copyright claim.”
In addition, users of ChatGPT could run into trouble for using content belonging to someone else to “feed” the program, which the technology uses to formulate the content.
‘Damn Be the Consequences’
Businesses using ChatGPT should take steps to protect their ownership in any content it produces, but the ability to so is sometimes “muddied up,” said Moish Peltz, who focuses on IP and emerging technologies at Falcon, Rappaport & Berkman in New York.
Lawyers counseling clients about legal pitfalls concerning the new technology sometimes find a high level of devotion to the platform, he said.
“Because of the magical and whimsical nature of ChatGPT, all of that normal due diligence that would go along with the use of an outside vendor kind of goes away. People are saying, ‘Wow! this is magic. I’m just going to use it and damn be the consequences,’” Peltz said. “I think maybe that works for some people, but I imagine it’s going to get some other people in trouble.”
Clients are experimenting with the new technology, drawn to its ability to turn tasks, such as writing a blog post or responding to an email, from a half-hour endeavor to one that can be knocked out in 60 seconds, Peltz said.
The latest version of the program was so much more powerful that the previous one that “it went from driving a Honda Civic to driving a Ferrari overnight. It’s really incredibly powerful,” Peltz said.
‘It’s Spitting Out Things’
Kendra Stephen finds some clients are so awestruck by ChatGPT that they trust it more than they should.
The Fort Lauderdale, Florida-based IP lawyer said her clients, whom she describes as young and entrepreneurial, are already making use of the new technology—and calling her with questions.
One client used ChatGPT to select a name for a new business, but that process did not include a check into whether the chosen name was in use by any other companies, she said.
“Although it can still be used as a tool, it’s not the end-all. It should be a starting point, but not the ending point,” Stephen said.
Stephen had another client who used ChatGPT to draft his own legal documents to be filed in court, but she warned him about the risks of that tactic.
“The issue with that is it’s not always correct. It’s spitting out things just to tell you things, but you have no way of verifying this, because you’re not an attorney and you don’t know how to do proper legal research,” Stephen said. “You could be putting together some random documents that you think will be held up in court. But it won’t, because it wasn’t properly drafted.”
‘Around to Protect You’
Some clients might decide the risks of using ChatGPT are so great that they tell their employees not to use it, or to use it only for producing internal content, said Wendi Opper Uzar, an IP attorney at Riker Danzig in Morristown, New Jersey.
Although the copyright office has opened the door to giving protection to certain works produced by AI, “that hasn’t actually occurred yet,” Uzar noted.
“It’s not clear what would meet that standard, and how we can make sure that our clients are meeting that standard so they can protect the works that they are creating,” Uzar said.
“It’s that balance of ‘can I use artificial intelligence’ and ‘can I tell my employees to use artificial intelligence?’ And if I do, do I have to give them guidelines and standards on how to use it, so we as a business can still protect our work that we’re putting out and make sure competitors and third parties aren’t going to be able to freely copy it? That’s the balance we’re working with,” Uzar said.
“There are many companies that are taking the position that they don’t want their employees using AI at all because of the potential implications it may have as far as their ownership in the material,” Uzar said.
For companies that choose to employ ChatGPT, Uzar advises them to check into the indemnification clauses of the company providing the AI platform, and to use only AI products sold by U.S. companies.
“If you use AI to generate content and you get charged with infringement, you want to make sure that company is going to be around to protect you,” she said.
IP lawyer Stephen, meanwhile, said that she encourages clients to use ChatGPT but not as an end-all resource. She said clients should use the new technology in conjunction with legal advice aimed at preventing misuse of the client’s content, and to ensure the client doesn’t infringe on content belonging to others.
“We do tell clients to use it. We don’t say don’t use it,” she said. “But we also say we want you to use it this way. You still want to have us review it.”
