Getty Images is suing Stability AI because the AI company trained its image generator on millions of Getty's copyrighted photos without permission or payment. A UK court already ruled the training infringed Getty's trademarks. The US case goes to mediation in October 2026, with a jury trial set for January 2028. Karma is a bitch.
## Why is the Getty Images lawsuit ironic?
For three decades, Getty Images built an empire on other people's photographs. The company sold licenses for thousands of dollars while paying the photographers who shot the images a fraction of that. Now an AI company has done to Getty exactly what Getty did to photographers, and suddenly unauthorized use is a tragedy worth suing over.
I have watched this from the losing side. I am a [celebrity photographer Los Angeles](https://www.dclee.net/) clients call when they want the real thing, and for years I lost work I should have won because I was not on Getty. Clients would rather pay Getty's hefty fees just to get on the wire service. I could not compete with a monopoly, so I built my own lane.
## How many gigs did I lose to Getty?
Countless. That is not an exaggeration for effect. Event after event, the call would come in, the budget would be real, and then the client would go with Getty because Getty meant the wire. The wire meant every outlet on earth would see the photos within minutes. One independent photographer, no matter how good, could not offer that distribution.
The wire was everything. In the age of print deadlines and broadcast slots, speed of distribution decided who got hired. Getty owned the pipes. Independent photographers owned their talent. Guess which one the client paid for.
So Getty got the gig. Getty sent a photographer. The client paid Getty thousands. And the photographer who actually showed up and shot the event got a few hundred bucks. That was the business model, and it worked beautifully for everyone except the person holding the camera.
## What did Getty pay photographers vs what clients paid?
Let me lay out the math plainly, because the numbers tell the whole story.
A typical corporate event: the client pays Getty somewhere between $3,000 and $5,000 for coverage. The photographer Getty sends to shoot it gets $300 to $400. Read that again. The person doing the actual work, the person with 27 years of experience reading a room and catching the moment, takes home roughly ten percent of what the client paid. Getty keeps the other ninety percent for owning the wire and the brand name.
At $300 for a full event, the math stopped making sense for me a long time ago. I would rather sit home than shoot under those terms. And I did sit home, many times, while Getty collected thousands off work I could have done better.
This is not ancient history or one bad deal. This was the standard arrangement, repeated thousands of times across the industry, for years. Getty's entire empire was built on this spread: charge the client thousands, pay the photographer hundreds, pocket the difference at scale.
## Is Getty right that AI training is theft?
Here is the part nobody wants to say out loud: yes. Getty is right.
Training an AI image generator on someone's copyrighted photographs without permission or payment is stealing. It does not matter whether the thief is a photo agency or a software company. The principle is identical. Someone made the image. Someone else used it to make money without asking. That is theft.
Getty just spent 30 years proving this point from the wrong side of it. The company that paid photographers $300 while charging clients $5,000 is now in court arguing that its own work deserves protection and compensation. The argument is correct. The messenger is the problem.
I know what it feels like to be on the receiving end. I once shot an editorial of a luxury condo building in downtown LA. Six months later I saw my pictures being used in an ad. I had a lawyer send a cease and desist. They kept using the pictures anyway. So I filed in small claims court, and I won. They ended up paying me. Most photographers never fight back. The system counts on that.
## What is the status of the Getty vs Stability AI lawsuit?
Two cases, two countries, both moving.
In the United Kingdom, Getty already won the first round. In November 2025, the High Court ruled that Stability AI's inclusion of Getty trademarks in AI-generated images constituted trademark infringement, and that the model provider, not the user, bears responsibility. The court also confirmed that Getty's copyrighted works were used to train Stable Diffusion. That ruling matters because it kills the industry's favorite defense: that the model is just math and nobody is responsible.
In the United States, the case is in the Northern District of California. Mediation was ordered for October 2026, and a jury trial is scheduled for January 2028. That is a long road, and AI companies know it. Lawsuits take years. Training took weeks. The incentives are obvious.
Watch what happens in mediation. If Getty settles for a licensing deal, every AI company will treat that as the price list. If it goes to a jury in 2028 and Getty wins big, the entire training-data economy gets rewritten.
## What should AI companies pay photographers?
Here is my position, and it is specific. AI companies should pay ongoing royalties when they use a photographer's images to create other images. Not a one-time buyout. Not an opt-out buried in terms of service. Royalties.
Think about how music works. When a song gets played, the songwriter gets paid, every time, through a system the industry built a century ago. Photography never got that system, which is exactly why Getty could pay $300 on a $5,000 job. But AI training is the perfect use case for it. Every image in a training set is identifiable. Every model knows what it trained on. The accounting is possible. The will is what is missing.
A royalty model would change the math for everyone. The [corporate photography](https://www.dclee.net/blog/corporate-photographer-los-angeles) I shoot for brands, the events covered at [rates that respect the work](https://www.dclee.net/corporate-event-photographer-los-angeles), all of it becomes training data the moment it hits the internet. If every use paid a fraction of a cent back to the photographer, the people who actually make the images would finally participate in the value they create.
The broader numbers back up the urgency. The global photographic services market was worth $37.51 billion in 2025 and is forecast to reach $39.21 billion in 2026, on its way to $48.91 billion by 2031, according to Mordor Intelligence. Meanwhile economist Joseph Politano found the US has lost more than 200,000 creative jobs in four years. The money is growing. The jobs are shrinking. That gap is where the stolen value goes.
## Who is the villain: the monopoly or the machine?
So here we are. Getty, the company that built a fortune paying photographers pennies, is suing Stability AI for doing the same thing to them. The thief is crying theft. And the worst part is that the thief is right.
I do not feel sorry for Getty. But I want them to win, because a win for Getty is a win for every photographer who ever got $300 on a $5,000 job. If the courts establish that training on copyrighted images without payment is infringement, that precedent protects all of us, including the people Getty spent decades underpaying.
That is the karma. Not that Getty suffers. That Getty has to argue our case to save itself.
So I will ask you the question I keep asking myself: who is the villain here? The monopoly that robbed photographers, or the machine robbing the monopoly?