For years, Meta trained its AI programs using the billions of public images uploaded by users onto Facebook and Instagram’s servers. Now, it’s also hoping to access the billions of images that users haven’t uploaded to those servers. Meta tells The Verge that it’s not currently training its AI models on those photos, but it would not answer our questions about whether it might do so in future, or what rights it will hold over your camera roll images.

On Friday, TechCrunch reported that Facebook users trying to post something on the Story feature have encountered pop-up messages asking if they’d like to opt into “cloud processing”, which would allow Facebook to “select media from your camera roll and upload it to our cloud on a regular basis”, to generate “ideas like collages, recaps, AI restyling or themes like birthdays or graduations.”

By allowing this feature, the message continues, users are agreeing to Meta AI terms, which allows their AI to analyze “media and facial features” of those unpublished photos, as well as the date said photos were taken, and the presence of other people or objects in them. You further grant Meta the right to “retain and use” that personal information.

Meta recently acknowledged that it scraped the data from all the content that’s been published on Facebook and Instagram since 2007 to train its generative AI models. Though the company stated that it’s only used public posts uploaded from adult users over the age of 18, it has long been vague about exactly what “public” entails, as well as what counted as an “adult user” in 2007.

Meta tells The Verge that, for now, it’s not training on your unpublished photos with this new feature. “[The Verge’s headline] implies we are currently training our AI models with these photos, which we aren’t. This test doesn’t use people’s photos to improve or train our AI models,” Meta public affairs manager Ryan Daniels tells The Verge.

Meta’s public stance is that the feature is “very early,” innocuous and entirely opt-in: “We’re exploring ways to make content sharing easier for people on Facebook by testing suggestions of ready-to-share and curated content from a person’s camera roll. These suggestions are opt-in only and only shown to you – unless you decide to share them – and can be turned off at any time. Camera roll media may be used to improve these suggestions, but are not used to improve AI models in this test,” reads a statement from Meta comms manager Maria Cubeta.

On its face, that might sound not altogether different from Google Photos, which similarly might suggest AI tweaks to your images after you opt into Google Gemini. But unlike Google, which explicitly states that it does not train generative AI models with personal data gleaned from Google Photos, Meta’s current AI usage terms, which have been in place since June 23, 2024, do not provide any clarity as to whether unpublished photos accessed through “cloud processing” are exempt from being used as training data — and Meta would not clear that up for us going forward.

And while Daniels and Cubeta tell The Verge that opting in only gives Meta permission to retrieve 30 days worth of your unpublished camera roll at a time, it appears that Meta is retaining some data longer than that. “Camera roll suggestions based on themes, such as pets, weddings and graduations, may include media that is older than 30 days,” Meta writes.

Thankfully, Facebook users do have an option to turn off camera roll cloud processing in their settings, which, once activated, will also start removing unpublished photos from the cloud after 30 days.

The feature suggests a new incursion into our previously private data, one that bypasses the point of friction known as conscientiously deciding to post a photo for public consumption. And according to Reddit posts found by TechCrunch, Meta’s already offering AI restyling suggestions on previously-uploaded photos, even if users hadn’t been aware of the feature: one user reported that Facebook had Studio Ghiblified her wedding photos without her knowledge.

Correction, June 27th: An earlier version of this story implied Meta was already training AI on these photos, but Meta now states that the current test does not yet do so. Also added statement and additional details from Meta.

Sendi Jia, a designer running her own studio between Beijing, China, and London, England, says she mainly uses AI generators like DALL-E to make fake photos for background panels or websites when her clients don’t have access to real ones. That’s helped clients with limited budgets, but it’s also exposed just how much of the creative process AI can replace. Recently, a potential client working in a university contacted Jia about creating the logo for a new project. Then, they changed their mind. They had used AI to make it, they said.

Chinese graphic artists are rapidly experiencing the impact of image generators on their day-to-day work: the technology enables copycats and profoundly shifts clients’ perception of their work, specifically in terms of how much that work costs and how much time it takes to produce. Freelance artists or designers working in industries with clients that invest in stylized, eye-catching graphics, like advertising, are particularly at risk.

Long before AI image generators became popular, graphic designers at major tech companies and in-house designers for large corporate clients were often instructed by managers to crib aesthetics from competitors or from social media, according to one employee at a major online shopping platform in China, who asked to remain anonymous for fear of retaliation from their employer.

Where a human would need to understand and reverse engineer a distinctive style to recreate it, AI image generators simply create randomized mutations of it. Often, the results will look like obvious copies and include errors, but other graphic designers can then edit them into a final product.

“I think it’d be easier to replace me if I didn’t embrace [AI],” the shopping platform employee says. Early on, as tools like Stable Diffusion and Midjourney became more popular, their colleagues who spoke English well were selected to study AI image generators to increase in-house expertise on how to write successful prompts and identify what types of tasks AI was useful for. Ultimately, it was useful for copying styles from popular artists that, in the past, would take more time to study.

“I think it forces both designers and clients to rethink the value of designers,” Jia says. “Is it just about producing a design? Or is it about consultation, creativity, strategy, direction, and aesthetic?”

At ad agencies, for example, graphic designers work on comprehensive strategies for campaigns, aiming to create iconic, recognizable visual identities across a variety of formats. As such, AI image generators are less useful because they don’t produce anything particularly unique, according to Erbing, a graphic designer in Beijing who has worked with several ad agencies and asked to be called by his nickname.

“Each project faces different problems, and designers are there to solve specific problems, not to create identical visuals,” he says. “Sometimes, the process of thinking through a project takes longer than actually creating the visuals.”

When faced with more complex tasks, AI’s utility dwindles. Image generators are capable of creating many images, but that does not replace the work of understanding what an ad campaign needs to establish a visual identity and communicate what it is the client is selling and why people should buy it. Then, translating those concepts to the AI productively is its own challenge. Among graphic designers in China, there’s a joke that using an AI image generator is like gacha, referring to addictive games where users spend money to receive randomized items and find out what they won.

“You might get a good result, but there will inevitably be dozens or even hundreds of poor ones,” Erbing says. “Personally, I see [AI image generators] as more of a toy than a tool.”

Across the board, though, artists and designers say that AI hype has negatively impacted clients’ view of their work’s value. Now, clients expect a graphic designer to produce work on a shorter timeframe and for less money, which also has its own averaging impact, lowering the ceiling for what designers can deliver. As clients lower budgets and squish timelines, the quality of the designers’ output decreases.

“There is now a significant misperception about the workload of designers,” Erbing says. “Some clients think that since AI must have improved efficiency, they can halve their budget.”

But this perception runs contrary to what designers spend the majority of their time doing, which is not necessarily just making any image, Erbing says.

Erbing, like other designers, hopes AI image generators can become more useful to graphic designers in the future, and notes that people’s perception of their usefulness outpaces their actual application. In the meanwhile, it is twisting the clients’ view of the usefulness of the artists themselves.