Shutterstock is reaping more profit for now—by training AI models

The image media library made $104 million last year

Shutterstock’s database was valuable enough for OpenAI to ink a six-year deal for scraping its image pool for training data. Google, Meta, Amazon and Apple followed suit. The latest deal involves Reka AI, a startup developing large language models.

What does this mean for the stock industry?

Stock has been predicted to wither and die as editors and publications opt for more customized generative AI versions of their database. For now, Shutterstock sees an opportunity here—even if it means hastening the stock photo industry’s own demise. Assuming the creators in their library have agreed to have their images scraped—and are being compensated accordingly—Shutterstock and its suppliers have a lot to gain by providing their images to AI companies, if only in the short term.

How is this deal different?

Like competitor agencies Getty Images and Adobe Stock, Shutterstock already provides an AI Generator so that users can create images trained on the database. However, in this case, the company is selling images to third-party giants for what’s estimated to become $250 million in revenue by 2027. It’s a giant paradigm shift in the way stock photo agencies are fighting for survival and using their enemy’s energy to defeat them, both for their own monetary gain and industry relevance.