OpenAI joins a growing list of companies that say they will protect their paying clients if copyright lawsuits are brought against them in relation to using OpenAI products. Well, sort of.
In a new blog post, the company says, "OpenAI is committed to protecting our customers with built-in copyright safeguards in our systems. Today, we’re going one step further and introducing Copyright Shield—we will now step in and defend our customers, and pay the costs incurred, if you face legal claims around copyright infringement. This applies to generally available features of ChatGPT Enterprise and our developer platform."
Other companies have also promised to indemnify paying customers against copyright and intellectual property claims. They include powerhouses like IBM, Microsoft, and Amazon.
But why are companies making this promise — and how real is it?
Why OpenAI Says It Will Pay For Customer's Legal Costs In Copyright Cases
Pretty much since the moment generative AI stepped onto the scene, artists and rights holders have screamed foul play. That's because major platforms like ChatGPT and Dall-e were "trained" using copyrighted material — but the people who own that source material aren't getting compensated, even though OpenAI is on pace to earn more than $1 billion annually from the software.
This matter extends to most forms of generative artificial intelligence. That's because the technology in principle isn't necessarily creating anything "new" via human authorship, it's just copying other material to generate an amalgam based of user prompts. Because algorithms trained on existing material don't pass the muster for human authorship, which also means they don't pass the muster to create their own copyrightable works. At least, that's one of the major arguments.
In reality, the whole situation is much less clear. We're in an area of relative unknown when it comes to the legality of these platforms. And the United States government is playing catchup. Even the Biden Administration's sweeping executive order, which was generally well-received by all sectors of the industry, made little mention of generative AI's potential infringing upon copyrights and instead focused on larger existential threats and privacy concerns.
Which means the field is wide open for a litany of litigation against the companies like OpenAI that are now making billions of dollars via products based on allegedly illegal use of works.
So, likely in an effort to not scare away potential and existing customers (well, customers who pay a lot of money), the company created its "Copyright Shield" system.
What Does OpenAI's Copyright Shield Actually Mean For Users?
We know very little about what OpenAI actually intends to do with its "Copyright Shield" program. Not surprisingly, one paragraph in a blog update isn't particularly satisfactory in covering the details of the program.
For instance, Copyright Shield certainly doesn't make any promises to help its customers actually defend the claims. Just pay for legal costs. Also, it only applies to enterprise customers and not free or basic paid plan users. For reference, an enterprise user of ChatGPT reportedly pays at least $60 per "seat" per month with a 150-seat minimum and a 12-month contract. Yeah, that's a minimum of $108,000 per year. So how many users does the Copyright Shield really protect?
Which of course begs a pretty major question around whether or not the company is committed to defending the core legality of its technology. Instead, it just seems kind of like a general insurance clause. Somewhat akin to a company making sure one of its vendors has a certain type of insurance before entering into a bigger business arrangement.
We also don't really know the likelihood of users going after a company that uses an OpenAI product versus OpenAI itself. Adobe is another company that offers its users some level of protection in case they're sued for copyright infringement by using Adobe products. But there's no real structure around what this protection actually entails, or if the company has any responsibility to actually stick to it.
OpenAI is currently facing a class action lawsuit from a group of authors including people like George R. R. Martin, the author of Game of Thrones. In a statement to ABC, OpenAI didn't really address the actual legal issues, but instead says it thinks creators would ultimately benefit from using ChatGPT. It's not really a defense, to say the least. And when a user can ask the platform to generate text in the style of a George R. R. Martin novel, you can make a pretty safe bet that ChatGPT isn't generating text on a complete whim.
So Should I Use Generative AI Products?
There are a lot of questions around the use of generative AI, particularly when it comes to legal and ethical concerns. Even if a court rules that it's legal for companies to train models on copyrighted works without compensating the original creators, is it ethical? After all, plenty of platforms have found an imperfect but still practical way to address use of copyrighted materials — like YouTube with music licensing.
However, it's unlikely that individual users will face direct copyright claims simply for using generative AI products. Though not impossible. It's kind of a "buyer beware" situation.
But as more and more creators bring their concerns around the technology to light, it's clear that there's a pretty big rift in the creative community. Is using this technology for general productivity as "bad" as using it to generate lyrics to a song in the style of Tupac that you then go on to record and make money from? And where do things like the protections Hollywood writers secured against AI come into play? If major industries are acknowledging the dubious origins of generative AI, will other companies start trying to compensate those whose works it uses to create products?
We have far more questions than answers. But you know we'll be paying close attention to all the developments.