There is much debate over whether the free exercise of artificial intelligence (AI) models on copyrighted material is permitted under the law. This debate is escalating to court cases filed in various jurisdictions and companies challenging companies that train their generative AI models by stealing copyrighted material without permission.
The debate is largely framed as an all-or-nothing proposition. AI model makers claim that all training is clearly permitted by law. Some even argue that there is no copyright protection for online content as pointed out by Mustafa Suleyman, CEO Microsoft AI, u remarks earlier this year. “I think that with the content that’s already on the open web, the social contract of that content since the 1990s is that it’s fair use. Anyone can copy it, recreate with it, reproduce with it. It was free software, if you will, it was an understanding,” Suleyman said.
As CEO of Getty Images, I am responsible for an organization that employs more than 1,700 individuals and represents the work of more than 600,000 journalists and creators worldwide. Copyright is at the very core of our business and the livelihood of those we employ and represent. As I’m sure you can imagine, I strongly disagree with the broad position outlined by Mr. Suleyman and others. I also doubt that Microsoft would vehemently disagree if the same logic were applied to their software and game titles.
The creative community, which represents a significant share of the global economy, mobilized against the unauthorized use of their shared text, music, photography and video. Recently, this mobilization was manifested in more than 30,000 artists who signed a statement that “the unlicensed use of creative works to train generative artificial intelligence is a major, unjust threat to the livelihoods of the people behind those works and should not be allowed.” I strongly agree with this statement. A world where artists cannot invest and be rewarded for their work is a world with less creativity and fewer people able to make a living from their trade. That is not a plausible future.
This disagreement underscores why we are litigating Stability AI in the US and the UK. We have not given Stability AI permission to use the millions of images owned and/or presented by Getty Images to train their Stable Diffusion model which is commercially available from August 2022 (Editor’s Note: Stability AI denies all allegations). The action against Stability AI alone represents a multi-million dollar investment and is expected to take years to resolve. However, identifying unauthorized use, securing evidence, and taking legal action against each individual company that trained on Stability’s content is prohibitively expensive, even for a company the size of Getty Images.
As the disputes slowly progress, AI companies are making the argument that there won’t be AI without the ability to freely collect training content, resulting in our inability to harness the promise of AI to solve cancer, mitigate global climate change, and eradicate global hunger. Keep in mind that companies investing in and building AI spend billions of dollars on talent, GPUs, and the necessary power to train and run these models — but incredibly, compensating content owners is an insurmountable challenge.
My focus is to achieve a world where creativity is celebrated and rewarded AND a world free of cancer, climate change and global hunger. I want my cake and eat it too. I guess most of us want the same.
We need to be open to a more nuanced discussion about AI and copyright. The reality is that there is a set path here: fair use (and similar concepts around the world) allow third parties to use copyrighted works without a license, subject to the following criteria:
- The purpose and character of the use, including whether such use is of a commercial nature or is in the public interest.
- The nature of the copyrighted work.
- The quantity and significance of the part used in relation to the copyrighted work as a whole.
- The effect of the use on the potential market or value of the copyrighted work.
Fair use applies on a case-by-case basis. So let’s not look at AI as one monolithic case, but break it down for what it really is: a wide range of models, possibilities, and potential applications.
Does cancer treatment affect the value of Kevin Bacon’s performance? Of course not. Does addressing climate change affect the value of Billie Eilish’s music? Of course not. Does solving global hunger affect the value of Stephen King’s writing? Again, clearly not. Not only does it not harm the value of their work, they would probably never object to such use if it could serve those purposes, even if such use might be commercial in nature. As the CEO of Getty Images, I can say that we would never debate or challenge these apps and would wholeheartedly welcome any support we could offer these apps.
But let’s now look at a more limited area of the AI universe: content generation models. These are models that generate music, photos, and videos based on text or other input. These models are trained on content from artists who do not have their permission, are commercialized by companies that target the same end markets that these artists make their living from, and clearly contradict the first and fourth criteria of fair use. Importantly, these models have no potential to raise our social outcomes. It is pure theft by one group for the financial benefit of another.
Fair use is a way to have our cake and eat it too. It has been established for a long time and we have to refuse invitations to ignore or override it. And that doesn’t mean content generation models can’t exist. On the contrary.
As licensed models Spotify and Apple Music has evolved from the original Napster infringing copyright, there are AI models developed with permission and business models that reward creators for their contributions. Like Apple Music and Spotify, they will cost a bit more, but they can thrive and be widely adopted if we create a level playing field by addressing those companies that choose to “move fast and break things”, in this case, infringing on established copyrights law.
There is a fair path that rewards creativity and fulfills the promise of artificial intelligence. Let’s stop the rhetoric that all AI training without permission is legal and that any demand to respect the rights of creators is to the detriment of AI as a technology.
More must-read comments posted by him Wealth:
Opinions expressed in Fortune.com comments are solely the views of their authors and do not necessarily reflect the opinions and beliefs Wealth.
