{"id":1460,"date":"2026-06-04T13:36:34","date_gmt":"2026-06-04T13:36:34","guid":{"rendered":"https:\/\/silvybrand.com\/?p=1460"},"modified":"2026-06-04T13:36:34","modified_gmt":"2026-06-04T13:36:34","slug":"let-us-filter-ai-slop-google-youtube-meta-instagram-tiktok","status":"publish","type":"post","link":"https:\/\/silvybrand.com\/?p=1460","title":{"rendered":"Let us filter AI slop, you cowards"},"content":{"rendered":"


\n<\/p>\n

\n
\n

It\u2019s almost impossible to avoid seeing AI-generated content online, but it doesn\u2019t have to be this way. YouTube, Instagram, TikTok, and more have ramped up content authentication efforts over the last year, with many now automatically applying labels to distinguish AI-generated images, videos, and music from those made by real, human creators.<\/p>\n<\/div>\n

\n

That\u2019s all very well and good if we\u2019re just stumbling across labeled content at random, but you know what would be better? Letting us filter out the AI slop.<\/p>\n<\/div>\n

\n

Current labeling efforts haven\u2019t meaningfully changed how content is presented online. You may notice that some TikTok<\/a> or YouTube videos<\/a> in your feeds now have AI disclosures in the description, or information labels overlaid onto the clip itself. Meta takes a similar approach by applying \u201cAI info\u201d labels to images on Facebook<\/a> and Instagram<\/a> that carry identifying AI metadata or voluntary disclosures from the creators.<\/p>\n<\/div>\n

\n

But if you want to actually avoid<\/em> seeing anything tagged with such labels \u2014 which is justifiable, given the brain rot it induces on top of the ethical and environmental concerns around generative AI \u2014 it\u2019s actually incredibly difficult to do so. A filter would easily solve this. All we need is an \u201cAI\u201d checkbox to toggle.<\/p>\n<\/div>\n

\n

I reached out to Meta, Google, TikTok, and Spotify to ask if they have plans to let users filter the various content they\u2019ve been authenticating with AI labeling systems. TikTok and Spotify never responded, and Google said it had nothing to share. Meta didn\u2019t provide an attributable comment. But to summarize, none of these companies said \u201cyes.\u201d<\/p>\n<\/div>\n

\n
\n
\n
\n
\n
\"A<\/a><\/div>\n<\/div>\n<\/div>\n

Notice how DeviantArt chose \u201csuppress\u201d here rather than \u201cexclude.\u201d<\/em><\/figcaption>Image: DeviantArt \/ The Verge<\/em><\/cite><\/p>\n<\/div>\n<\/div>\n<\/div>\n
\n

One of the only online platforms that I\u2019ve seen with an AI content filter is DeviantArt, and its implementation is extremely telling. For one, you can\u2019t access it on DeviantArt\u2019s feeds or store page, so it feels somewhat hidden away. Instead, you have to make an account and then hover over your user icon at the top-right of the page to find the \u201cAI Content Settings\u201d menu. From there, you only have two options: the default \u201cShow AI\u201d setting, or the \u201cSuppress AI\u201d option that claims you\u2019ll see \u201cfewer instances\u201d of AI-generated or manipulated imagery.<\/p>\n<\/div>\n

\n

Having tried both options, I, unfortunately, don\u2019t see a notable difference. I\u2019ve got a pretty good eye for spotting AI-generated \u201cdigital illustrations\u201d at this point, but I didn\u2019t have to rely on my suspicions alone \u2014 almost every dubious image I selected included a creator\u2019s disclosure in the description that confirmed the work was spat out by a robot. DeviantArt does a poor job of automatically applying AI labels<\/a> to images with metadata that clearly indicates AI provenance.<\/p>\n<\/div>\n

\n

Pinterest<\/a> has a similar system in place. Users who are signed into a Pinterest account can click on the settings icon, select \u201cRefine your recommendations,\u201d and then tap the \u201cAI content\u201d tab to toggle specific categories, including art, beauty, fashion, and home decor. Disabling any of these options will show you \u201cless AI-modified content\u201d for that particular category, according to Pinterest, but in my experience, it\u2019s far from perfect. The setting is also arguably harder to find than a filter built into Pinterest\u2019s feeds. I still saw plenty of images with suspicious AI tells (including uncannily perfect photography models and unexplainable illustration errors), despite the AI filters being maxed out.<\/p>\n<\/div>\n

\n
\n
\n
\n
\n
\"A<\/a><\/div>\n<\/div>\n<\/div>\n

I commend the customization options here, but these refinement options are hidden away and don\u2019t actually work effectively.<\/em><\/figcaption>Image: Pinterest \/ The Verge<\/em><\/cite><\/p>\n<\/div>\n<\/div>\n<\/div>\n
\n

And that is almost certainly what will happen if other platforms like YouTube or Instagram introduce an AI content filter: It won\u2019t work very well. But that\u2019s okay because it would expose the ineffective \u201csolutions\u201d our AI emperors dress themselves in. They exist, on paper, to appease regulators and critics, but do little to address the actual problem of distinguishing AI fakery from authentic photography and creative works.<\/p>\n<\/div>\n

\n

And platforms do<\/em> know it\u2019s a problem. Instagram head Adam Mosseri said in December<\/a> that \u201cauthenticity is becoming a scarce resource\u201d amid the rise in AI-generated content. And now we have Google CEO Sundar Pichai admitting<\/a> in a recent Decoder<\/em> interview that \u201cthere\u2019s a lot of AI slop out there,\u201d and that online users need to \u201cadapt to it.\u201d Okay, give us filters.<\/p>\n<\/div>\n

\n

Provenance-based systems like C2PA and SynthID work by embedding metadata or invisible watermarks into content at the point of creation. But there are plenty of open-source AI models that don\u2019t do this (especially if they\u2019re designed for nefarious purposes), and even then, metadata can be stripped out too easily to make this dependable. There are also detection-based methods that analyze patterns in digital content and then rate the likelihood that AI was used to create it, but these can provide false positives. None of this currently works effectively at scale.<\/p>\n<\/div>\n

\n

Nevertheless, companies, including AI providers like OpenAI<\/a>, are currently heralding those AI labeling solutions as something that will help prevent people from being duped by deepfakes and other misleading fakery. If regulators caught wind of how ineffective they are, then online platforms and AI providers may need to actually find a solution that does<\/em> work, instead of what currently feels like a smokescreen.<\/p>\n<\/div>\n

\n

Platforms will argue that they risk incorrectly flagging authentic content if they push labeling initiatives too hard. Both Meta<\/a> and YouTube<\/a> found out the hard way after applying AI labels to images and videos that creators said were produced without the help of such tools. If that\u2019s such a concern for current labeling systems, then find a better solution<\/em>. Surely improving the user experience for your millions of users is a worthwhile investment to fend off competition?<\/p>\n<\/div>\n

\n

And while I\u2019m asking, why can\u2019t I report all the unlabeled AI slop I see every day? Given the scale of the issue \u2014 with a study by Kapwing last year<\/a> finding that more than 20 percent of YouTube videos shown to new users is low-quality generated slop, for example \u2014 I imagine a lot of human moderators would be required to effectively vet each report.<\/p>\n<\/div>\n

\n

And perhaps that\u2019s the rub. At a time when big tech is replacing workers with AI that can supposedly outperform them<\/a>, can it afford to backtrack on its carefully constructed narrative by hiring them back to fix AI\u2019s problems? Humans tend to have pesky requirements like salaries and benefits<\/em>, compared to automated moderation systems that lack nuanced investigation skills.<\/p>\n<\/div>\n

\n

An alternative to labeling AI-generated content would be to start labeling verified human creators instead<\/a>. That wouldn\u2019t necessarily identify synthetic content posted by<\/em> those creators, but it could help us to see less from unverified content farms that churn out low-quality slop. This is the future that Instagram\u2019s Mosseri pitched<\/a> for Meta\u2019s image-sharing platform and something that Spotify is already doing with Verified artists<\/a>.<\/p>\n<\/div>\n

\n
\n
\n
\n
\n
\"A<\/a><\/div>\n<\/div>\n<\/div>\n

This sounds great! If only there was a way to exclusively surface music by verified artists\u2026<\/em><\/figcaption>Image: Spotify<\/cite><\/p>\n<\/div>\n<\/div>\n<\/div>\n
\n

Of course, Meta, Spotify, and Google don\u2019t just host AI-generated imagery, ads, and music; they\u2019re also responsible for making the tools that create it. That\u2019s why they insist that not all<\/em> AI content is slop and that it\u2019s more an issue of quality \u2014 if it gets convincing enough, they\u2019re hoping you won\u2019t notice and keep happily slurping from the trough. Allowing users to filter it out regardless would go against all the effort these platforms have undertaken to profit from AI: They want<\/em> you to embrace the slop factory.<\/p>\n<\/div>\n

\n

I\u2019m happy to be proven wrong. I\u2019m actually outright pleading<\/em> for online platforms to prove that AI labeling efforts haven\u2019t been a waste of time. But right now, they hold all the cards and we\u2019re left to just hope that their AI moderation efforts are up to snuff. So give us a basic \u201cno AI\u201d or \u201cverified human creator\u201d filter and we\u2019ll be the judge of how well this is actually going.<\/p>\n<\/div>\n

Follow topics and authors<\/strong> from this story to see more like this in your personalized homepage feed and to receive email updates.<\/span><\/p>\n