I have a folder of photos that have been viewed 77 million times and downloaded over 300,000 times. I made zero euros from any of it. And for years, that felt fine to me. Lately it doesn’t feel quite as simple.
Let me explain what I mean, because the easy version of this story is “naive photographer gives away work, regrets it later.” That’s not what happened. The harder, more honest version is that the deal I thought I signed up for quietly changed underneath me, and I didn’t notice until I started paying attention to where stock photography actually goes once you hit upload.
Why I Did It in the First Place
I’ve been making images long before Planbnb or Microfilming existed. The camera came first. The businesses came later, partly because of it. So when I started putting photos on a free stock platform, the motivation was never financial. It was exposure, it was a portfolio that lived outside my own websites, and honestly, it was the simple satisfaction of knowing a photo I took in Kavala might end up in someone’s project on the other side of the world.
That trade made sense to me. I give you a free, high-quality photo. You use it in a blog, a presentation, a small business website that could never afford a photographer. I get visibility and the quiet ego boost of big numbers next to my name. Nobody loses.
That was the deal as I understood it in 2014, 2017, even a few years ago.
The Deal Changed Without a New Contract
Here’s what I missed for a long time: stock platforms don’t just sit on a library of free photos waiting for individual humans to download them one at a time anymore. Across the industry, libraries have been licensed in bulk to AI companies training image generation models, and the photographers whose work makes up those libraries have mostly found out after the fact.
77 million views taught me that giving things away can be one of the most generous things you do as a creator.
I post exclusively on Pexels, so I went and actually read its terms instead of skimming them like I did the first time. Pexels is owned by Canva, and its terms state directly that content I upload can be used by Pexels, its affiliates, and any third party partners it authorizes to develop, improve, and provide their services, including through machine learning, without providing me attribution or compensation. Pexels’ own AI and ML FAQ confirms this covers things like Canva’s generative AI products and systems built to detect AI-generated content.
There is an opt-out. It’s sitting in account settings, and I never knew it existed until I went looking for it this year.
I never agreed to that specific use when I first uploaded my work. I agreed to a terms of service document I skimmed once, years ago, that contained language broad enough to cover something I hadn’t thought through. That’s not the same as informed consent. It’s the same trick every platform on the internet uses: get permission once, in the broadest possible language, then expand what that permission quietly means later without asking again.
Three Ways I’ve Tried to Think About This Honestly
I don’t think the answer here is outrage, and I don’t think it’s naive acceptance either. A few lenses have helped me actually work through it instead of just having a feeling about it.
The consent lens. When I uploaded those photos, “free” meant free for a human being to use in their project. It did not mean “free raw material for a commercial AI system that might compete with photographers for the same work.” Those are different transactions, even if one piece of legal text was stretched to cover both. I gave consent for the first. I never meaningfully gave it for the second.
The value lens. Somebody is making real money from those 77 million views and 300,000 downloads. It’s not me. It’s Pexels and Canva, and increasingly, whatever AI products get built on top of that library. The value I created didn’t disappear, it just got captured somewhere else in the chain. That’s worth sitting with, because it reframes “I did this for free” into “I did this for free, and a billion-dollar design company is now building products off the back of it.”
The agency lens. Would I upload the same photos today, knowing what I know now? Some, yes, because exposure still has real value to me and I’m not interested in becoming bitter about a system I voluntarily joined. Others, no, particularly anything that feels personal or distinctly mine rather than generic stock material. The honest answer isn’t “stop giving anything away.” It’s “stop giving things away on autopilot.”
What I Actually Changed
I stopped treating “free” as a single category. Now I separate my work into two buckets before I ever upload anything: material I’m genuinely happy to see used by anyone, for any purpose, including training a model, because it’s generic enough that I have no real attachment to its downstream use; and material that carries something more specific to me, my eye, my location, my way of seeing Kavala, that I keep out of platforms with broad AI licensing terms.
I also went back and checked the opt-out setting on Pexels directly. It’s added after the backlash against platforms quietly licensing libraries for AI training, buried in account settings nobody reads twice. It took me ten minutes to find it and switch what I wanted switched.
None of this makes me anti-AI or anti-stock-photography. I still believe in the original trade: free work in exchange for visibility and the genuine pleasure of being useful to a stranger’s project. What I no longer believe is that “free” is a fixed, simple word. It’s a contract that can be quietly rewritten by parties who aren’t you, and the only real protection is paying attention to what you’re actually agreeing to, not what you think you agreed to five years ago.
77 million views taught me that giving things away can be one of the most generous things you do as a creator. It also taught me that generosity without attention is just an open door, and you don’t always get to choose who walks through it.