Why are Getty and Shutterstock on opposite sides of the AI legal debate?

For many people, whether they work in an enterprise business or write their own indie blog, Getty Images and Shutterstock are similar companies, synonymous with the world of stock photos. Different prices, different offerings, but still — both offer stock imagery to add to digital or print content.

But over the past few months, it has become clear that the two companies have diverged in their efforts to deal with the exploding landscape of AI-powered text-to-image generation, and the quickly-evolving legal issues. It is, it seems, a perfect example of what Michael Eshaghian, an attorney at the Los Angeles-based Mesh IP Law, calls the “growing pains” of this new AI technology “until we settle into a legal equilibrium.”

Getty Images sues Stability AI, Shutterstock aligns with Meta

For example, today, the Verge reported that Getty Images intends to sue Stability AI, the creators of the open-source text-to-image generator Stable Diffusion, in the U.K.

Getty claims that Stability AI ‘unlawfully’ scraped millions of images from its site (the full suit is not public and Stability AI said they haven’t yet received it).

On the other hand, just a few days ago Shutterstock announced it was expanding its relationship with Meta to “use its datasets to develop, train and evaluate its machine learning capabilities.” This followed the company’s announcement in October that it was partnering with OpenAI to integrate DALL-E 2 into its offerings, with plans to offer compensation to artists — and OpenAI CEO Sam Altman revealed that the firm licensed imagery from Shutterstock to train DALL-E beginning in 2021.

At that time, Getty Images CEO said in an interview, “I think we’re watching some organizations and individuals and companies being reckless […] I think the fact that these questions are not being addressed is the issue here. In some case, they’re just being thrown to the wayside. I think that’s dangerous. I don’t think it’s responsible. I think it could be illegal.”

In a press statement around the Stability AI lawsuit, Getty Images said that it “believes artificial intelligence has the potential to stimulate creative endeavors. Accordingly, Getty Images provided licenses to leading technology innovators for purposes related to training artificial intelligence systems in a manner that respects personal and intellectual property rights. Stability AI did not seek any such license from Getty Images and instead, we believe, chose to ignore viable licensing options and long-standing legal protections in pursuit of their standalone commercial interests.”

Why the drastically different approaches?

Eshaghian pointed out that the last major revision in U.S. copyright law was in 1976, well before the Internet, let alone the current generative AI revolution.

“As with any hugely disruptive technology, the law often lags behind, and when this happens you’re going to see different parties approach the technology differently, as we now see with Shutterstock and Getty’s diametrically opposed approaches,” he told VentureBeat.

Getty, he said, has a reputation of “being aggressive with their copyrights” and added that he “wouldn’t be surprised if Getty moves forward with their lawsuit.”

Not all agree, however.  Legal scholar Andres Guadamuz, a reader in intellectual property law at the University of Sussex in the UK who has been studying legal issues around generative AI, said that the Getty lawsuit is interesting because “it signals that Getty wants a licensing agreement with Stability, much like the one between Shutterstock and OpenAI. Getty knows the future is AI, they want a piece of the action.”

And Bradford Newman, who leads the machine learning and AI practice of global law firm Baker McKenzie, maintains that while Getty has compared the current legal landscape of generative AI to the early days of digital music and Napster, the analogy does not hold.

“Unlike in the music business, there are not large publishing houses and similar existing organizations with whom these AI developers can enter into licensing or royalty agreements,” he said. “It is impractical for the developers to have to enter into tens of millions of individual agreements with the originator of each image scraped.”

Carrot vs. stick options

Newman said he assumes — but hasn’t checked — that Shutterstock’s Terms of Service allows them to own the IP rights and/or license them to third parties, which the Meta partnership press release appears to support.

“It’s a proverbial win-win,” he said. “Meta gets access to millions of images to train its AI, and through its agreement with Shutterstock, erases any potential for lawsuits like the Getty one and those to come.  And Shutterstock realizes revenue from its deal with FB and pays its artists for their contributions to training AI models.

Jim Flynn, managing director of law firm Epstein Becker Green, pointed out that at each stage of technological evolution, content owners face a choice.

“Some choose carrot and some choose stick,” he told VentureBeat. “We saw it with music, and then with books, and now we are seeing it with AI software that, while generating new content, is also dependent in many cases on existing content.”

Flynn said he is interested in watching how the dueling parties, and courts, ultimately address the fair use and other questions “as each of those sides in AI-related litigations swing sticks that each may have some heft legally.” But the Meta/Shutterstock deal, he said, is the most interesting:

“They seem to have decided to limbo together beneath the sticks they could have been swinging at each other,” he said. “It will be interesting to see how many decide to avoid, dare I say ‘slip under,’  the litigation obstacles in the same way.  If the market rewards them, others may follow that strategy.”