There is much debate over whether it is legal to freely train artificial intelligence (AI) models on copyrighted material under the law. This debate is reaching court filings in various jurisdictions and companies challenging companies that train their creative AI models by infringing copyrighted materials without permission.
The debate is largely framed as an all-or-nothing proposition. AI model makers say all training is legal under the law. Some claim that there is no copyright protection for online content, as the company’s CEO Mustafa Suleyman pointed out. Microsoft AI, inside remarks made at the beginning of the year. “I think that with regard to the contents that are already on the open network, since the nineties the social contract of these contents is fair use. Anyone can copy it, reproduce it, reproduce it. That was free, if you will, that was the understanding,” said Suleyman.
As CEO of Getty Images, I oversee an organization that employs more than 1,700 people and represents the work of more than 600,000 journalists and creators worldwide. Copyright is at the heart of our business and the livelihood of those we employ and represent. As I am sure you can imagine, I do not agree with the broad position expressed by Mr. Suleyman and others. I suspect Microsoft would disagree if the same logic were applied to their own software and game titles.
The creative community, which represents a large part of the world economy, has mobilized against unauthorized use of their collective text, music, photography and video. Recently, this mobilization showed more than 30,000 artists signing a statement “The unlicensed use of creative works to train creative AI is a huge and unfair threat to the livelihoods of the people behind those works, and should not be tolerated.” I strongly agree with this statement. A world where artists cannot be invested in and rewarded for their work is a world with less creativity and less people able to make a living at their craft. That is not a plausible future.
This discrepancy underscores why we are pursuing lawsuits against Stability AI in the US and the UK. We have not given Stability AI permission to use millions of images owned and/or represented by Getty Images to train its Stable Diffusion model. of 2022 (Author’s note: Stability AI denies all claims). Action against stability AI involves a multi-billion dollar investment and is expected to take years. However, identifying unauthorized use, securing evidence, and pursuing litigation against all companies trained on Stability content is prohibitively expensive, even for a company of Getty Images’ scale.
As the lawsuits slowly progress, AI companies present an argument that there will be no AI without the ability to freely harvest training content, and as a result we cannot take advantage of AI’s promise to solve cancer, mitigate global climate change, and eradicate world hunger. Note that companies investing in and building AI spend billions of dollars on talent, GPUs, and the power required to train and run these models, but they vehemently claim that compensation for content owners is an insurmountable challenge.
My focus is on achieving a world where creativity is celebrated and rewarded AND a world free of cancer, climate change and hunger. I want to have cake and eat it too. I suspect most of us want the same.
We need to be open to a more nuanced discussion of AI and copyright. The reality is that there is an established path here: fair use (and similar concepts around the world) can be taken from copyrighted works without permission by third parties, based on the following criteria:
- The purpose and nature of the use, including whether the use is commercial or in the public interest.
- Nature of the copyrighted work.
- The amount and prominence of the portion used in relation to the entire copyrighted work.
- 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 look at AI not as a monolithic case, but instead explore what it really is: a wide range of models, capabilities, and potential applications.
Does curing cancer affect the value of Kevin Bacon’s performances? Obviously not. Does addressing climate change affect the value of Billie Eilish’s music? Obviously not. Does solving world hunger affect the value of Stephen King’s writing? Again, clearly no. In addition to not harming the value of their work, they would probably never question such use if it served those purposes, even if such use could be of a commercial nature. As the CEO of Getty Images, I can tell you that we will never discuss or question these apps and would very much welcome any support we can provide for these apps.
But let’s now look at a more limited area of the AI universe: content creation models. These are models that generate music, photos, and videos based on text or other input. These models are being trained on artists’ content without permission, commercialized by companies targeting the same end markets that those artists live in, and clearly run afoul of the first and fourth criteria of fair use. It is important to note that these models do not have the power to improve the results of our society. It is pure theft from one group for the financial benefit of another.
Fair use is the way to have our cake and eat it too. It has been established for a long time and we have to ignore the calls to reject or cancel it. And that doesn’t mean content creation models can’t exist. On the contrary
As a licensed model Spotify and the apple Evolved from the original infringer Napster are AI models developed with music, 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 become widely accepted if we create a level playing field by targeting companies that choose to “move fast and break things”, in this case breaking established copyright. the law
There is a fair way that rewards creativity and delivers on the promises of AI. Let’s stop the rhetoric that all unauthorized AI creations are legal and that any demand to respect the rights of creators is at the detriment of AI technology.
More comments posted by this one should be read luck:
Opinions expressed in Fortune.com comments are solely the opinions of the authors and do not necessarily reflect their views and beliefs. luck.
