VSCO Updates Its Policy to Explicitly State It Won’t Train AI on Your Photos

In May, VSCO came under scrutiny for wording in its Terms of Use that seemed to give the company permission to use certain user content to develop, train, and improve generative AI or machine learning models. It has now taken the extra step to clarify that it doesn’t do that, and won’t.

It was noticed that VSCO’s Terms of Use stated that users granted VSCO the right to use the “name, image, voice, or likeness of any individual in your Content, in whole or in part, and in any form, media, or technology, whether now known or developed in the future.” It also included “the right” for VSCO to “use certain Content for Creator Promotion and to develop, train, and improve AI or machine learning models as further described in our Creator Content Standards.”

This, of course, set off some alarm bells and the company quickly said it wouldn’t be training AI models on users’ content. For some, this might not have been enough, however, and VSCO seems to have taken note of that.

“You deserve a straight answer about what happens to the work you put on VSCO,” Eric Wittman, CEO of VSCO, writes in a letter sent to VSCO users today. “Sharing your work shouldn’t mean wondering whether it will be used to train a system designed to generate content (photos or video) without you.”

As a result, the company has today updated its Creator Content Standards to more explicitly state this stance.

“We don’t use your content to train generative AI models. Not your public work. Not your private work. Not because you have a paid membership. The same protection applies whether you pay us or not,” Wittman continues.

“We’re updating our Creator Content Standards to make that explicit.”

The updated Content Creator Standards page now says the following:

If you are a paid member with an active VSCO membership or a non-paying Creator on VSCO, your Public or Private Creator Content will not be utilized to develop, train, or improve AI or machine learning models for Generative AI Use.

“For the past 15 years we have championed photographers. That means building a business around helping you make your work and earn from it. It doesn’t mean treating the photographs you entrust to us as raw material for models that generate images, video or other media. It’s a good moment to also be precise about what we mean by generative AI because AI isn’t one thing, and calling out the distinction matters,” he writes.

“We use machine learning technology and non-generative AI across the VSCO platform with the goal of helping photographers, whether that’s to find work, search for people to connect with, edit images and get more out of our products. We have heard from the photography community this is where AI is useful. That’s very different from using your photographs to train models that generate new content. Our updated standards make that boundary explicit.”

VSCO says that it isn’t just holding itself to these standards, but anyone the company works with, too.

“We’re holding our providers to that distinction, too. Outside AI providers who have been powering VSCO features have already agreed not to use content submitted through our services to train generative AI models. Language in our previous Creator Content Standards was too broad to make these distinctions and intentions clear. That left room for confusion about where we stood. You shouldn’t have to decipher a policy to find out what is happening with your work on a platform you use,” Wittman says.

“Your work isn’t just a file on our platform. It represents the time you spent learning, the places you went to, the people who trusted you and the choices only you could make. That perspective and what you choose to do with it is yours.”

VSCO’s updated Creator Content Standards can be read on the company’s website.