Can you make money selling AI photos and videos on stock image sites?

After AI books on Amazon KDP, we reach the format where the flood of AI-generated content isn't a future risk — it's a market collapse that has already happened and can be measured in concrete numbers. Of this entire cluster, it's the clearest case of what happens when AI content supply grows without limit against demand that doesn't grow at the same pace.

The market collapse: the number that explains everything

Combined total revenue for stock photographers fell from USD 1.47 billion in 2019 to USD 31 million in 2026 — a 98% drop. The cause is visible in the leading platform's catalog: Adobe Stock's library is now roughly 48% AI-generated, with about 29 million new AI images uploaded every month. Content supply exploded exactly when per-image license demand didn't — the result is the worst possible scenario for anyone entering this business expecting the numbers from a few years ago.

What each platform pays for AI content (and who bans it outright)

Platforms split into completely different AI policies. Adobe Stock explicitly accepts it (mandatory checkbox on upload, metadata scanning, C2PA signatures) and pays between USD 0.25 and 0.50 per download of AI content — somewhat less than the USD 0.33-1.00 range for traditional content, within a 33%-of-net-sale scheme with a USD 0.33-0.38 floor per license. Shutterstock heavily restricted uploads of AI generated with external tools, requiring mandatory metadata disclosure. iStock only sells AI content generated with its own in-house engine. Getty Images, following its legal dispute with Stability AI, flatly doesn't accept any content primarily generated by AI (Midjourney, DALL-E, Stable Diffusion) — it only admits traditional photography with technical AI enhancements (noise reduction, resolution upscaling, minor retouching).

How much you can actually make today (industry figures, unaudited)

Industry sources (not independently verified) estimate a serious AI stock photographer can reach USD 200-2,000 a month by 6 months in, and USD 2,000-10,000 a month by 18-24 months, with a catalog of 5,000 or more commercially-tuned images. Those figures don't carry the verification level of other cases in this cluster — worth treating as an optimistic industry ceiling, not an expected average, especially given the sector's 98% total-revenue collapse.

What still works amid the saturation

The strategy reported by operators still generating income is twofold: use Adobe Firefly (Adobe's native generator, with a better acceptance rate on Adobe Stock since it comes from the same ecosystem) and target specific commercial niches instead of generic images — exactly the opposite of uploading indiscriminate volume, which is what caused the price collapse in the first place.

So, is it worth it?

It's the most extreme example in this entire cluster of what happens when the marginal cost of producing AI content falls to near zero while the market buying it doesn't grow: a 98% revenue drop for the entire sector between 2019 and 2026. It's worth entering only with expectations adjusted to that reality, checking each platform's policy first (Getty rules this path out entirely) and targeting specific commercial niches instead of competing on volume against millions of new images every month.