Pinterest Begins Labeling Generative AI Content

Good news for those who are sick of their Pinterest feed being inundated with AI-generated trash: Pinterest has now officially implemented a new policy to label AI-generated images.

Earlier this month, Futurism published a report which highlighted the rising influx of AI-generated content in the app, which is now overwhelming some users.

The report found that AI-generated “slop”, often used to redirect Pin users to spammy sites, has become so overwhelming that some have even stopped using the app as a result.

In response, Pinterest noted that it is working on new AI detection and labeling features.

As per Pinterest:

“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.”

Then last week, within its official documentation, Pinterest added this new proviso in its rules:

“Pinterest may display a label in the foreground of an image Pin when we detect that it has been generated or modified with AI. This is in accordance with IPTC standard for photo metadata. We’re working on ways to expand our capabilities to better identify GenAI content in the future through additional technologies.”

The IPTC Photo Metadata Standard is primarily used for alt-text descriptions, but it also includes info on image editing processes and tools used in the image’s creation. As a result, Pinterest can use this to highlight AI generated content based on the data of the image.

The new tags have been visible in the app for the last couple of months, on a seemingly small collection of Pins.

Here’s an example of what the tag looks like on desktop:

As you can see, the Pinterest “AI generated” or “AI modified” tags are overlaid at the bottom left of the image, so they’re not watermarks, but they do stand out, which should help to highlight AI-created content.

So yes, Pinterest is moving to better tag and highlight AI-generated images within the Pin feed, in order to help users avoid clicking through on these junk peddlers.

Though the same won’t extend to paid promotion:

“For ads on Pinterest, we only display a disclosure in the “Why am I seeing this ad?” section, indicating whether the ad includes modified or AI-generated content, as detected through metadata.”

So ads will be less transparent, but even so, Pinterest is moving to expand its detection and transparency measures, which, ideally, will have an impact on this element.

Though the influx of AI-generated content hasn’t seemingly impacted Pinterest usage in any major way as yet, with the company reporting record-high usage in Q4, with 553 million monthly active users.

Still, it could become a bigger problem, and it’s good to see Pinterest taking definitive measures to combat AI slop in the app.

Also worth noting: Pinterest has also added a new toggle which enables you to opt out of having your Pins used to train its AI models. Currently, it's opted in by default, with all user content (except those who've switched it off) eligible to be used for AI training from April 30th.