TikTok Experiments With Letting Users See Less AI Slop

TikTok has built its empire on one simple promise: open the app and the For You feed will know what you want before you do.

But lately, that promise has become a little more complicated.

As generative AI tools become faster, cheaper and easier to use, social platforms are being flooded with synthetic videos, fake clips, AI voiceovers, automated storytelling, low-effort animations and endless recycled content. In internet shorthand, a lot of it gets called “AI slop” — content that may be technically impressive at first glance, but quickly becomes repetitive, shallow or simply annoying when it takes over the feed.

Now TikTok is experimenting with a way to let users dial some of it down.

TikTok’s New AI Content Control

TikTok has been testing a new setting that gives users more control over how much AI-generated content appears in their For You feed.

The feature sits inside TikTok’s “Manage Topics” area, where users can already adjust how often they see certain types of content, such as sport, food, dance or current affairs. The new AI-generated content control works in a similar way. Rather than banning AI content outright, it gives users a slider-style preference tool that tells TikTok whether they want to see more or less of it.

That distinction matters.

TikTok is not offering a full “remove all AI” button. It is giving users a stronger preference signal. If someone enjoys AI history videos, synthetic art, fantasy edits or AI-generated explainers, they can still see more of that content. If someone feels their feed has become overrun with fake voices, uncanny videos and low-effort engagement bait, they can ask the algorithm to reduce it.

In other words, TikTok is trying to make AI content adjustable rather than unavoidable.

Why AI Slop Became a Problem

AI-generated content is not automatically bad. Plenty of creators use AI in smart, creative and genuinely useful ways. It can help with editing, captions, translation, ideation, visual effects and accessibility. For small creators and brands, AI can lower production costs and make content creation less intimidating.

The problem is volume and quality.

Because AI tools can generate videos at scale, the incentive is obvious. If one synthetic video can be made in minutes, hundreds can be pumped out in a day. Some accounts are built almost entirely around this model: grab a trend, generate a script, add an AI voice, create a strange visual, post repeatedly and see what the algorithm rewards.

That creates a feed-level problem. Users may not object to one AI video. They object to seeing the same style of content again and again. The fake podcast clip. The AI baby voice. The invented historical scene. The motivational monologue over stock footage. The obviously synthetic celebrity clip. The engagement-bait “you won’t believe this happened” story that did not happen.

Eventually, the feed starts to feel less human.

That is dangerous for TikTok because the platform depends on emotional immediacy. People open TikTok because it feels alive, chaotic, funny, intimate and current. If the For You page starts to feel like a machine talking to itself, users may scroll less, trust less and spend less time engaging.

A Sign That User Control Matters More Now

The new AI preference tool is also part of a wider shift in social media. For years, platforms trained users to accept algorithmic feeds as something mysterious and untouchable. You watched, liked, skipped, shared and commented. The algorithm interpreted your behaviour.

But AI content has made that system feel more fragile.

When synthetic content becomes too easy to mass-produce, platforms need better ways to understand what users actually want. A watch-time signal alone is not always enough. Someone might watch an AI-generated video because it is strange, irritating or confusing — not because they want ten more like it.

That is why explicit controls matter. A “show me less AI-generated content” option gives TikTok a cleaner signal than passive behaviour alone. It lets users say, directly, that a category of content is becoming too much.

For creators and marketers, this is important. Growth on TikTok has never just been about posting more. It has been about understanding what real users want to keep seeing. Tools and services such as Celebian that offer gradually delivered real views may help with social visibility, but long-term performance still depends on content that feels worth watching, sharing and trusting.

The Labelling Challenge

Reducing AI content only works if TikTok can reliably identify it.

That is not simple.

Some creators label AI content clearly. Others do not. Some AI-generated videos are edited, downloaded, reuploaded or stitched into new formats, which can strip away metadata or make detection harder. A clip may also be partly AI-generated rather than fully synthetic, which creates grey areas.

TikTok has been working on AI labels for some time, including labels for content made or significantly edited with AI. It has also supported industry standards such as Content Credentials, which can carry information about how a piece of media was created.

The newer development is invisible watermarking. TikTok has said it is testing invisible watermarks on AI-generated content made with its own tools and on uploads that include certain content credentials. The idea is to create a hidden signal that TikTok can read, making it harder for AI labels to disappear when content is edited or reuploaded.

This is not a perfect solution. Watermarking and detection systems can become a cat-and-mouse game. But it does show that TikTok sees AI identification as central to the future of feed quality.

What This Means for Creators

For creators, the message is not “never use AI.”

The message is: do not let AI become the whole point.

If users begin actively reducing AI-generated content in their feeds, lazy AI posts may struggle more. Content that looks mass-produced, emotionally empty or obviously templated could become easier for users to avoid. That does not mean every AI-assisted post will be punished, but it does raise the bar.

Creators who use AI well should focus on adding human value around it. That might mean stronger opinions, original storytelling, personal experience, real footage, better editing, authentic commentary or genuinely useful information.

AI can support the content. It should not replace the creator’s taste, judgement or personality.

This is especially true for brands. A brand that fills TikTok with generic AI clips may get short-term volume, but it risks looking cheap, detached or untrustworthy. A brand that uses AI to speed up production while still sounding human has a much better chance of building an audience that actually cares.

What This Means for Users

For users, TikTok’s experiment is a small but meaningful step toward a more customisable feed.

The For You page has always been powerful because it learns quickly. But that power can also make users feel trapped. Watch the wrong thing for a second too long and suddenly the feed changes. Engage with a trend once and the algorithm assumes you want more. Get pulled into a weird AI niche and it can feel like the app has decided your taste for you.

A visible AI content control gives people more say.

It will not fix every problem. It will not remove every fake clip. It will not magically separate good AI from bad AI. But it acknowledges something important: users do not all feel the same way about synthetic content.

Some people love it. Some people hate it. Most people probably sit somewhere in the middle, happy to see creative AI occasionally but tired of low-quality spam.

The Bigger Picture

TikTok’s experiment points to a bigger question facing every major platform: how much synthetic content is too much?

AI-generated media is not going away. It will become more realistic, more personalised and easier to produce. The platforms that win will not be the ones that simply allow infinite AI content into every feed. They will be the ones that can separate useful, creative AI from repetitive slop — and give users enough control to shape their own experience.

TikTok’s new AI slider is not a dramatic revolution. It is more like an early pressure valve.

But it matters because it shows the platform understands the risk. If users feel overwhelmed by AI-generated content, they may not blame the creators. They may blame TikTok itself.

The For You feed only works when it still feels like it is for you. If it becomes a dumping ground for synthetic filler, users will want a way out.

TikTok is now testing whether a simple “see less” control can help keep the feed human enough to stay addictive.