- within Intellectual Property topic(s)
- with Senior Company Executives and HR
- in United States
- with readers working within the Banking & Credit, Technology and Media & Information industries
As more companies turn to artificial intelligence for content for a wide range of uses, case law is starting to take place to ensure companies aren't using copyrighted material for profit.
While it's still moving through the process, it's important to understand you could be breaking the law without knowing it when you use AI for content.
Many of the initial court cases deal with how AI companies are using their source material. And the first rounds of rulings on fair use of materials have been mixed.
Let's start with how AI companies are building the database AI uses to write content for prompts.
In reality, AI scours all types of public sources quicker than ever before, taking that information to produce content, whether it is emails, blogs, internet codes, contracts and more.
With AI, the more information the technology receives, the better the outputs will be. That's why many of the companies that provide AI can use sources to "train" the computer for better responses.
So far, the input in the training has been the subject of court cases.
In one case, Kadrey v. Meta, the technology juggernaut Meta was accused of using copyrighted materials from 13 authors to train its "Llama" AI program. The judge ruled in favor of Meta, basically saying the output from Meta was transformative, not just regurgitation. In addition, the output was not just a copy of the author's work, but new copy.
That said, the judge did say Meta's use of the authors' works could dilute the market for their products and that could make the use of the materials illegal.
In Thomson Reuters v. ROSS, the owner of Westlaw accused ROSS of using the materials to create a competing product for Westlaw – a database of legal information for lawyers.
The judge ruled the output was not transformative and created a direct competitor for the original work. As such, the judge ruled in favor of Thomson Reuters and that ROSS illegally used copyrighted work.
Most businesses are not training AI to be better. They are using AI for content, so these suits would not apply directly.
However, at some point, someone will likely sue a business that uses an AI program found to have violated copyright laws. They will claim that the copyright use hurts their market share.
While this is still unknown legal territory, the U.S. Congress is starting to weigh in. In a July Legal Sidebar article, it said, "If a generative AI output infringes a copyright in an existing work, both the AI user and the AI company could potentially be liable under current law."
It will be difficult for a company to prove, but as such, we're suggesting only using AI to guide original content, not using AI to write for the business. If you do use content from a third-party AI provider, like OpenAI or Meta, have our firm review your contract and terms of agreement.
The attorneys at The Orlando Law Group help with all types of legal issues for businesses and technology in Orlando, Waterford Lakes, Altamonte Springs, Winter Garden, Lake Nona, St. Cloud, Kissimmee, and throughout Central Florida.
If you have questions about anything discussed in this article or other legal matters, give our office a call at 407-512-4394 or fill out our online contact form to schedule a consultation to discuss your case. We have an office conveniently located at 12301 Lake Underhill Rd, Suite 213, Orlando, FL 32828, as well as offices in Seminole, Osceola and West Orange counties to assist you.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
