Pinterest Implements AI Content Labeling and Data Usage Policies Amid User Concerns

Pinterest to Train AI Models on User Content

While Pinterest may not be an obvious platform for photographers, there are still plenty of them on there. For those shooters, they may want to take note of the recent news that the boards-based app will train AI models with user content. Pinterest's privacy policy will be updated on April 30, 2025. An opt-out option is available for those who do not want their work fed into a generative AI model. The new clause reads that Pinterest may use information "to train, develop and improve our technology such as our machine learning models, regardless of when Pins were posted." That means all content ever posted on the site is fair game. So if you have a long-forgotten account on Pinterest from when you were posting photos there circa 2010, then you might want to look into this if you don't want your work used for training. However, the company tells Futurism that the new update is codifying actions the company was already taking. "Nothing has changed about our use of user data to train Pinterest Canvas, our GenAI model," Pinterest tells the online news outlet. "Users can easily opt out of this use of their data by adjusting their profile settings." Pinterest is following in the footsteps of Meta, Reddit, and Google, all of which train AI models on their user's data. The overwhelming need for quality training data in the AI arms race has meant that large companies view such practices as "fair use" -- despite no definitive legal ruling on the hot topic. Meanwhile, Pinterest is rolling out AI-generated labels on posts created by, or with the aid of, generative AI tools. Social Media Today reports that the move is a response to the large amount of "AI slop" on the site. "We have been building labeling of AI-generated or modified content to provide relevant context to users about what they see on Pinterest. We'll continue to expand these labels in the coming months," says Pinterest, via Social Media Today. Some users already flagged the labels which appear as a small watermark in the bottom right-hand corner. Example here. Last month, Futurism ran an article titled "Pinterest Is Being Strangled by AI Slop" claiming users are being overwhelmed by the amount of AI content on the app.

Pinterest will no longer let AI content go unlabeled

Pinterest has implemented a new policy to label AI-generated images in response to user concerns about the increasing prevalence of such content in their feeds. This initiative aims to provide relevant context about the images users encounter on the platform. A recent report by Futurism highlighted that users have expressed frustration over the influx of AI-generated "slop," which often directs them to spammy websites, leading some individuals to abandon the app altogether. In acknowledgment of this issue, Pinterest stated, "We have been building labeling of AI-generated or modified content to provide relevant context to users about what they see on Pinterest. We'll continue to expand these labels in the coming months." Pinterest's official documentation now includes a provision indicating that the platform may display a label on an image Pin when it detects AI generation or modification. This labeling adheres to the International Press Telecommunications Council (IPTC) standard for photo metadata, which encompasses information on image editing and creation tools. The tags will appear at the bottom left of images, clearly indicating if a Pin is "AI generated" or "AI modified." Although these labels are designed to help users avoid junk content, they will not extend to paid promotions. For ads, Pinterest will only disclose in the "Why am I seeing this ad?" section whether the ad includes modified or AI-generated content, based on metadata analysis. Despite user concerns, Pinterest reported record-high engagement in Q4 2023, with 553 million monthly active users. However, the company acknowledges that the issue of AI content may develop into a bigger problem. As part of its ongoing improvements, Pinterest has also introduced a toggle feature that allows users to opt out of having their Pins used to train AI models, although the default setting is to allow this. Artists Google to court over AI image generator The rise of AI-generated content on the platform has drawn criticism from users who argue that it dilutes the quality and usability of Pinterest. Searches for common themes like recipes or DIY projects often yield numerous AI-generated results, many of which link back to AI-driven sites offering low-value content. Users lament that they can no longer find genuine human-made examples when searching for inspiration. Individuals creating AI content for financial gain, such as digital marketer Jesse Cunningham, promote strategies to generate income through AI-generated posts on Pinterest. Cunningham claims that his approach, which involves using AI images and text, has significantly increased traffic to his linked websites. Despite the criticism, Pinterest has not restricted AI-generated content but instead has released new AI tools aimed at helping marketers create ads more efficiently. In response to these developments, Pinterest emphasized that impressions from generative AI content account for a small percentage of total impressions, highlighting its commitment to enhancing user control over visible content. As users continue to express their need for greater control over AI presence in their feeds, many are turning to alternative platforms that offer filtering options to manage AI content visibility. Some users have noted that they find it increasingly challenging to discover authentic art and creative works amidst the flood of AI-generated images.