When Getty Images announced its display partnership with OpenAI, I didn’t only see an AI headline.
I saw a stock photography story – and I’ve lived one of those before.
Early in my career, I helped launch and scale Fotolia as a major contender in the newly-emerging microstock photography market. Back then, microstock felt like a revolution. It made professional‑quality images accessible to small businesses and independent creators who could never have afforded legacy stock. It turned talented amateurs into meaningful earners. It helped fuel an entire wave of web design, blogging, and digital marketing.
I was proud of that work. I still am.
But as the customer base shifted – from webmasters and small businesses to global brands and top‑tier publishers – I also watched a quieter cost emerge. “Good enough” imagery became good enough for everyone, including major corporate clients who once paid a premium to master stock artists. The same mechanisms that democratized access also compressed value at the top.
So when I look at Getty’s partnership with OpenAI, I don’t just see an integration.
I see the next chapter of a story that started long before AI.
Microstock Was the First Great “Creative Compression”
Before microstock, stock photography was expensive, gated, and dominated by a few large agencies. Big brands and big budgets could justify the fees. Small businesses, not so much.
Microstock flipped that:
- It opened the door to a vastly larger buyer base.
- It rewarded contributors who understood volume, keywording, and marketplace dynamics.
- It trained buyers to expect fast, searchable, affordable imagery on demand.
Access exploded. So did usage. A new generation of creators got a seat at the table.
But the economics changed just as dramatically:
- Single‑image value fell as libraries grew and prices dropped.
- Legacy pricing structures came under pressure.
- The perceived value of an individual image shifted from “crafted asset” to “commodity input.”
Microstock didn’t kill stock photography.
It rewired it. It moved the industry from scarcity to abundance, and from premium per‑image value to high‑volume marketplaces.
AI is about to run that play again – this time at a much larger scale.
Getty + OpenAI: Licensing vs. Training Are Two Different Fights
The Getty / OpenAI announcement matters because it pulls licensed imagery directly into AI‑powered search and discovery.
That’s the optimistic read:
- Licensed archives showing up in AI interfaces is qualitatively better than unlicensed scraping.
- It suggests at least some platforms accept that visual content requires real agreements, not just vague “fair use” arguments.
- It could create new discovery paths and revenue streams for high‑quality imagery.
Display, by itself, is not the problem.
In fact, display done right is part of the solution.
The real tension lives elsewhere: in training.
Displaying licensed content inside an AI interface and using that content to train or improve generative models are two very different things:
- Display can preserve attribution, support licensing, and send value back into an existing ecosystem.
- Training can enable systems to synthesize near‑substitutes at scale, often without ongoing participation from the people whose work made that synthesis possible.
That raises hard questions that look suspiciously familiar to anyone from the stock world:
- Was my work included?
- On what terms?
- Was there a license or opt‑in?
- Can the system now generate market substitutes for what I sell?
- If so, do I participate in that value – or am I just part of the training set?
Those aren’t philosophical questions.
They’re business, labor, and rights questions.
We ignored some of those questions in the early microstock years. We don’t get to ignore them with AI.
The Pattern: We Keep Trading Creative Economics for Scale
Microstock taught the industry a powerful but uncomfortable lesson:
When you maximize access and scale, you also rewrite who gets paid, how much, and for what.
AI is set up to push that pattern to its limit.
With microstock, buyers shifted from “Can I afford an image?” to “Which of these thousands will I license?”
With AI, buyers may shift from “Which image should I license?” to “Why license anything if I can generate a custom visual on demand?”
The potential outcomes are mixed:
- Licensed archives like Getty may gain new distribution and monetization channels inside AI.
- High‑quality photography may stand out as a mark of credibility in a world flooded with synthetic imagery.
- New licensing constructs – rights‑managed training sets, contributor royalty pools, verified “human‑made” marks – could emerge.
But there’s a darker version:
- Generative systems trained on vast libraries reduce demand for licensed stills.
- Style imitation and synthetic substitution undercut the earning potential of working photographers.
- The economic value of a lifetime of creative work gets absorbed into infrastructure, not income.
Microstock compressed the value of many individual images.
AI risks compressing the value of entire careers.
Getty’s Role Isn’t Neutral
Getty is not a passive player in this story.
It sits at the intersection of:
- Huge archives (editorial, commercial, historical, cultural).
- Deep rights and licensing infrastructure.
- Legal willingness to challenge AI companies when lines are crossed.
That combination gives Getty leverage that individual creators, small agencies, and independent archives simply do not have.
So the question isn’t just:
- “Is this a smart deal for Getty?”
It’s also:
- “Does this deal set a precedent that strengthens or weakens the connection between creative work and creative compensation?”
- “Do photographers and illustrators participate in the upside of AI products built on top of licensed libraries?”
- “Are we building a model where rights and attribution matter, or one where content is just fuel for someone else’s engine?”
A licensing deal that only improves one platform’s UX and one company’s P&L is a narrow win.
A licensing model that keeps creative contributors in the value chain is a different kind of win entirely.
What Brands and Marketers Should Pay Attention To
If you commission, license, publish, or rely on images for your brand, this isn’t background noise. It’s your operating environment.
Four practical questions are worth tracking:
- Where does AI sit in your image discovery workflow?
If your teams start using AI interfaces as the first stop for research and inspiration, you need clarity on what’s licensed, what’s synthetic, and what’s in between. - How will you differentiate licensed vs. generated visuals?
There will be contexts where AI‑generated images are fine – even useful. There will also be contexts where they’re a reputational risk. Editorial, healthcare, nonprofit, political, and news‑adjacent work will not be neutral territory. - What’s your standard for rights inside AI tools?
“It came out of the interface” is not a legal strategy. Brands will need clear guidance on when an image is truly licensed, what indemnification looks like, and how they document that in a mixed human/AI workflow. - How will you continue to fund original visual reporting and creativity?
If every workflow gradually tilts toward “good enough and cheap,” the pipeline of people doing the hard, expensive visual work shrinks. That’s not just an ethical problem; it’s a strategic one. Brands that care about trust, nuance, and long‑term equity will still need human eyes and human judgment.
This Time, We Know How the Story Can Go
When microstock showed up, most of us were busy building. We saw the opportunity much more clearly than the second‑order effects.
We don’t have that excuse anymore.
- We know that expanding access changes economics.
- We know that marketplaces can centralize power even as they democratize participation.
- We know that “just a new channel” can evolve into a new norm that is hard to unwind.
AI will not end photography. It will, however, change:
- How images are discovered.
- How they’re created.
- How they’re priced.
- How they’re credited.
- Who participates in the value they generate.
The Getty / OpenAI partnership is not the end of that story. It’s an early, visible sign of how the next phase might be structured.
If we care about preserving a healthy creative ecosystem – one where talent can still build a livelihood, where archives are valued, and where “real” still means something – we should be asking hard questions now, not after the economics have quietly shifted.
Microstock taught us that democratizing access can expand an industry while quietly rewriting its incentives.
AI has the power to do both on a far larger scale.
The real thought‑leadership opportunity is not to declare this partnership good or bad in absolute terms, but to insist on a standard:
If AI becomes the new front door for visual discovery, the people whose work fills those rooms deserve more than a line item in someone else’s training data.
That, more than the headline, is what’s at stake in deals like Getty and OpenAI.
