Is this against TikTok's rules?

The Community Guidelines, under integrity and authenticity, prohibit attempts to artificially increase engagement or to deceive the recommendation system. Named directly: bots and automated accounts; buying followers, likes, views or comments; the trade or marketing of services that sell them; and publishing instructions on how to artificially increase engagement.

The consequences are account-level. Fake engagement is one of the deceptive behaviours for which TikTok says it may restrict an account's ability to post, to appear in search or to appear in the For You feed — the state most creators mean by the word "shadowban".

The word doing the work in every line is "artificial". The rule is written around automation and money, because those are the two things that produce engagement without a person wanting to give it.

The For You feed Eligibility Standards are broader than the rules, and they govern recommendation rather than removal. A video can break no rule and still be judged unsuitable to push to a broad audience, and when it is, it stays up and reaches almost nobody new.

Two categories matter here. Engagement bait — content whose purpose is to extract likes, comments, saves or shares, "comment X for Y" and its relatives — and false incentives to follow, including promises to follow back. Both are reasons a video is not recommended, with nothing to appeal because nothing was flagged as wrong.

Neither category is about who watched your video. Both are about what the video itself asks of the viewer.

No money changes hands for engagement. Points are earned by engaging and cannot be bought, so there is no trade of the kind the rule names.

Nothing is automated. Every engagement is a person opening the TikTok app, and the flow measures the time they spent there from the server rather than from a countdown on the page. There are no bots because there is nothing for a bot to run.

Nothing is held. No password is ever asked for, no official API is used, no token is stored. Ownership is proved by leaving 5 real comments, which is work only the person holding the phone can do.

Nothing is asked of your videos. Matching sends people to your video; it puts no bait in it, and it never asks anyone to follow — a follow exchange was assessed and deliberately not built.

And it is still coordination. A real person choosing to watch a video in their own subject is not a bot, and the rule draws its line at "artificial". Whether TikTok's systems draw that line in the same place for paced, matched, human attention is not published, and no service can tell you it is. That is the honest gap, and every limit on the safety-rails page is there to keep the pattern inside what a real audience would produce.

PodSwap's own risk assessment rated six possible TikTok actions for the chance of a participant-visible penalty within twelve months. The follow exchange carried the highest figure of the six, at 30–50%, and it is the one that was not built. The actions that were built are the ones the assessment rated lower, and they were then wrapped in the daily caps, the randomised gap, the delivery ceiling and the action-mix limits — all of which are published with the configuration key that sets them.

That is the most anyone can honestly say. A service that tells you there is no risk is describing something it cannot know.