AI Image Generators Are Replacing Stock Photos

I’m a guy on both sides of the image business. As a stock photographer, I spent several years creating perfect shots of isolated subjects with natural soft shadows, good lighting, and sharp details on a white background. As a customer, I bought the images I needed for clients’ websites and my own online projects. I still buy stock images occasionally, perhaps once every six months.

Most stock photo agencies are incredibly disrespectful toward their photographers and contributors. They also have a remarkable appetite for money, especially when somebody else did the work.

The Photographer’s View

Some stock agencies keep 85% of a sale while the non-exclusive photographer receives 15%. I have personally earned as little as $0.06 from a licensed image. Yes, six cents. That is not a royalty. That is the sound of a coin falling behind a sofa.

An iStock/Getty contributor reported receiving numerous royalties of just $0.01 for images licensed in China. Can it get any lower?

I once felt proud to have a portfolio of more than 2,000 images. Ten thousand would have been better, of course, because the traditional solution to miserable stock photography earnings is to produce an even larger amount of underpaid work. However, deleting 2,000 images becomes much easier when AI is taking over stock photography and can generate high-quality, realistic images within seconds.

The decision to delete my portfolios was not easy. I had to consider all the time I had spent shooting and doing post-production. Each image represented an idea, preparation, photography, editing, keywording, and uploading. But respecting the time I had already invested did not mean I had to continue selling my work for six cents. In the end, the Delete button was the right choice.

First, I deleted my portfolios from Dreamstime, 123RF, and Depositphotos. Shutterstock used to be my best seller until I started seeing $0.10 royalties far too frequently.

Shutterstock, especially as one of the largest stock agencies, could afford to pay more. Instead, it invented a system that gives it another chance to pay contributors less at the beginning of every year. Contributors work their way up through royalty levels, only to be reset to the lowest level in January. It is a loyalty program in which the company collects your points.

Well then …

What kept me there was the naive hope that Shutterstock might eventually get better. It did not. Hope is sometimes a lovely thing. In business, it can also be an unpaid waiting room.

For now, I still have an online portfolio with Adobe Stock because it is currently my best seller, and I can usually hit the payout button every two weeks. I may delete my other remaining portfolio, on iStock, next.

Stock agencies do not seem to see stock photographers as human beings who create images with passion. They do not respect the time behind each photograph. The photographer pays for the camera, lenses, lights, props, software, electricity, storage, travel, and sometimes models or locations. The photographer also carries the risk that an entire shoot will produce nothing useful.

Then comes the unpaid administrative work. Images must be selected, edited, described, keyworded, uploaded, categorized, and submitted for review. A reviewer may reject an image for a reason another reviewer completely ignored the day before. Even an accepted image can disappear beneath millions of competing files after an algorithm changes.

The agency controls the prices, subscriptions, search results, royalty calculations, and payout threshold. The contributor controls the exciting opportunity to upload more.

A visible “Thank you” after an upload does not pay the bills.

Inventing new ways to pay photographers less instead of finding better ways to sell their work is sick. Stock agencies chose the lazy solution: push prices down, offer cheap subscriptions, and let contributors absorb the difference. Who cares about the people creating the product? Before AI, without photographers, those agencies would have had nothing to sell.

Perhaps that is why they are so excited about AI. It creates inventory without expecting royalties, recognition, or basic respect.

Stock Photographers Are Guilty Too

Part of the responsibility lies with photographers. This is not only a story about greedy stock agencies. Some people will produce hundreds or thousands of images and accept six-cent sales without doing anything about it. They tell themselves that every little sale counts and that it will eventually add up.

Six cents does add up. So does dust.

By continuing to upload more work under these conditions, contributors help keep royalties exactly where they are. Agencies have no reason to improve when photographers continue supplying fresh material regardless of how little they receive.

Contributors have discussion forums where they complain about tiny royalties, falling sales, changed algorithms, and unfair systems. What they cannot do is make a collective decision to remove their portfolios. Many see another contributor leaving and think, “Good. More sales for me.”

That may produce a few extra downloads in the short term, but it also tells the agency that another photographer can always be replaced. One leaves, ten remain, and twenty more arrive after watching a video promising passive income from stock photography.

Photographers also flood the agencies with endless variations of the same subjects. More handshakes. More salads. More attractive people laughing at laptops. More isolated objects photographed from nine nearly identical angles. Everyone uploads more because individual images sell less, and then individual images sell even less because everyone uploaded more.

Stock photographers may be the only workforce where a pay cut makes half the room angry while the other half opens Lightroom and uploads another 300 files.

The agencies understand this perfectly. Contributors are spread across the world, have different living costs, and compete against one another. There will always be somebody willing to accept less. As long as photographers treat every departing contributor as extra shelf space for themselves, there will never be enough pressure on agencies to change.

Complaining while continuing to provide the same amount of work is not a protest. It is feedback with free inventory attached.

The Buyer’s View: AI Does It Better

For years, I bought photos from stock agencies for clients’ websites and my own projects. I still do when I find the right image, but these days that happens perhaps once every six months.

What wore me down was the endless supply of smiling people. Apparently, every manager, doctor, mechanic, and call center employee spends the entire working day grinning at objects.

Finding a real photo of a serious-looking person took time. Finding the right composition and pose took more time. Searching for the right image was often exhausting.

Today, I can ask AI something like this:

Generate an image of a serious-looking manager in a modern office. Use natural lighting and soft shadows. Keep realistic skin texture and ordinary imperfections.

Then I can continue working with the result. If I do not like that the manager has both hands on the table, no problem. I can request a different pose and get it within a minute. If I need more empty space on the left for text, I ask for it. If I want a different office, camera angle, expression, shirt, or lighting, I ask for that too.

AI is not perfect. It can still generate strange fingers, unreadable text, and furniture that appears to have been designed during a fever. But correcting or regenerating an image is often faster than searching through hundreds of stock photos that are almost right.

Some AI image generators are free, while others are paid. Either way, the value and convenience can be much greater than paying a stock photo agency for an existing image that I cannot properly customize.

As a photographer, I can say, “It sucks.” As a buyer, I can say, “That’s awesome.” Both are true.

Stock photo agencies see the potential too. They also see another opportunity to reduce their dependence on contributors and keep more of the profit.

As of September 3, 2026, Shutterstock displays this message: “Skip the scrolling. Instantly generate one-of-a-kind images with our AI generator.”

Well, why should I use Shutterstock’s AI generator? Why not use the tools people already use, such as ChatGPT and Grok? Shutterstock is effectively advertising that searching through its enormous stock library is an inconvenience and that customers should generate something instead. It spent years building the haystack and has now opened a business selling magnets.

It is only a matter of time before some agencies close, merge, or keep their brands alive with far fewer contributors and employees. They may find ways to survive through AI tools, legally verified content, premium human photography, or genuine editorial images. Those areas still have value, but they will never generate the enormous revenue that mainstream commercial stock photography once did.

I may still find and buy a useful stock image occasionally. But for generic subjects, why search through a giant catalog when AI can create the exact photograph, illustration, or increasingly even the video footage that I need? Sometimes I need it for work. Sometimes I simply want it for fun. Either way, the old catalog suddenly feels very old.

AI Has Taken Over

I think AI marks the beginning of the end for much of conventional stock photography. Not all photography, of course. Real news, sports, celebrities, historical events, and documentary evidence still require real cameras and real photographers. AI cannot honestly document something that actually happened.

But generic commercial stock photography is different. A serious manager in an office, a family eating breakfast, or a smiling customer-service employee does not need to be a specific real person. It only needs to look convincing and fit the project.

Technology keeps moving. In some fields, there is a meaningful improvement every month or even every week. AI becomes more capable, realistic, controllable, and human-looking. Its mistakes are gradually disappearing, while the reasons to search through traditional stock libraries are disappearing with them.

People have predicted that AI is just a bubble waiting to burst since at least 2023. Gary Marcus, an American cognitive scientist and professor emeritus of psychology and neural science at New York University, predicted that the generative AI bubble would collapse in 2025. In August 2024, he went further and suggested that it might happen within “days or weeks.” That particular countdown has now been running for a remarkably long collection of days and weeks.

Of course, an investment bubble can burst without the technology disappearing. Companies can fail, ridiculous valuations can collapse, and investors can lose money while people continue using useful AI tools. The dot-com bubble burst too. Unfortunately for anyone waiting for the Internet to go away, it remained annoyingly available.

No, I am not a scientist. But I can tell you that AI will not disappear while people want tools that can generate, check, and do work for them. I will always want AI that checks my grammar or generates the featured image above this article. And I’m too lazy to set up my camera, stand in front of it, take the photo, transfer it to my computer, and edit it in Photoshop. It is faster to upload a few reference photos of my face and body to an AI image generator and ask it to “add a camera with an attached flash.” I would rather use that time to write this article.

Yeah, something old ends and something new begins. But I have to admit that taking photos for stock agencies was genuinely exciting. My images appeared on book covers, flyers, websites, and in magazines. I loved seeing my work out there.

I also made a substantial financial investment in cameras, lenses, lighting, software, and other equipment. More importantly, I invested years of my time.

For what? Six cents per image, even before AI image generators existed? It sucks.