[dentemple]

> After a “small number of clicks” the researchers encountered pornographic content ranging from women flashing to penetrative sex.

I (40m) don't think I've ever seen literal flashing or literal porn on TikTok, and my algorithm does like to throw in thirst content between my usual hobby stuff.

Are they making the claim that showing porn is a normal behavior for TikTok's algorithm overall, or are they saying that this is something that specifically pervasive with child accounts?

[mothballed]

Does TikTok direct what you see based on what other accounts you interact with are interested in? I would expect teenagers to have a different interest profile than your average 40 year old. I would expect algorithms to more or less unwittingly direct you to the kind of stuff your peers were interested in.

[netruk44]

TikTok’s recommendations are based off as much info as it can get, really.

Approximate location, age, mobile OS/browser, your contacts, which TikTok links you open, who generated the links you open, TikTok search history, how long it takes you to swipe to the next video on the for you page, etc.

I don’t think it's really possible to say what TikTok’s algorithm does “naturally”. There’s so many influencing factors to it. (Beyond the promoted posts and ads which people pay TikTok to put in your face)

If you sign up to TikTok with an Android and tell it you’re 16, you’re gonna get recommended what the other 16 year olds with Androids in your nearby area (based on IP address) are watching.

[gadders]

Yeah, I've not seen any actual porn either. Just thirst traps.

It might be because I always block anyone with an OF link in their bio, but then that policy doesn't work on Insta.

[ivape]

You think thirst traps are okay for kids? If we rewind time, the Girls Gone Wild commercial is not supposed to be even remotely possible on certain channels.

We’re a derelict society that has become numb, “it’s just a thirst trap”.

We’re in the later innings of a hyper-sexualized society.

Why it’s bad:

1) You shift male puberty into overdrive

2) You continue warping young female concepts of lewdness and body image, effectively “undefining” it (lewdness? What is lewdness?).

3) You also continue warping male concepts of body image

[mvieira38]

This content isn't as overt as it may seem, maybe you did come across it and just didn't notice flashing. Those "in the know", generally younger people whose friends told them about flashtok, know what to look for

[causal]

Also: kids click on links adult ignore without thinking. Our brains have built in filters for avoiding content we don't want; for kids everything is novel.

[saurik]

I wonder when this study happened? FWIW, there was some pretty intense bombing of full-on nudity content to TikTok a month or two ago--it all looked like very automated bot accounts that were suddenly posting scenes with fully nude content cut out of movies--that I saw a number of people surprised were showing up in their feeds. It felt... weaponized? (And it did not last long at all, FWIW: TikTok figured it out. But it was intense and... confusing?)

[fsckboy]

>Are they making the claim that showing porn is a normal behavior for TikTok's algorithm overall, or are they saying that this is something that specifically pervasive with child accounts?

the latter is what they tested, but they didn't say specifically pervasive.

you quote the article so it seems like you looked at it, but questions you are curious/skeptical about are things they talked about in the opening paragraphs. it's fine to be skeptical, but they explain their methodology and it is different than the experience you are relying on:

>Global Witness set up fake accounts using a 13-year-old’s birth date and turned on the video app’s “restricted mode”, which limits exposure to “sexually suggestive” content.

>Researchers found TikTok suggested sexualised and explicit search terms to seven test accounts that were created on clean phones with no search history.

>The terms suggested under the “you may like” feature included “very very rude skimpy outfits” and “very rude babes” – and then escalated to terms such as “hardcore pawn [sic] clips”. For three of the accounts the sexualised searches were suggested immediately.*

>After a “small number of clicks” the researchers encountered pornographic content ranging from women flashing to penetrative sex. Global Witness said the content attempted to evade moderation, usually by showing the clip within an innocuous picture or video. For one account the process took two clicks after logging on: one click on the search bar and then one on the suggested search.

[NoGravitas]

Yeah, I (50m) have never encountered literal porn on TikTok. Suggestive stuff, thirst traps, sex ed, sex jokes, yes, but no literal porn or even nudity.

[Hizonner]

They are saying that they can find some of that content when they use relatively sophisticated techniques to intentionally try to find it.

They are in the business of whipping up outrage, and should not be given any oxygen.

[lupusreal]

> relatively sophisticated techniques

Clicking on thirst trap videos?

[thehodge]

Agreed, I've never even seen boobs on TikTok...

[elevation]

A soccer mom I know shared that she once tried TikTok. Within seconds of installing the app, the algorithm was showing nsfw content. She uninstalled it.

I assume that the offending content was popular but hadn’t been flagged yet and that the algorithm was just measuring her interest in a trending theme; it seems like it would be bad for business to intentionally run off mainstream users like that.

[yapyap]

after reading some of the article it seems to me that they’re saying that on a restricted account thats got the bday of a 13 year old with the suggested search terms tiktok shows and a few clicks you can see actual porn.

[ChromaticPanic]

I'm on an unrestricted account and I can't find actual porn. Sounds like this article is rage bait, claiming women in swim wear as porn.

[IanCal]

Really? I've signed up to bluesky and tiktok and on both have seen literal porn extremely early without engaging directly (such as liking or responding, speed of scrolling could be something).

[InitialLastName]

All of these apps are 100% using your scroll speed/how long you spend engaging with the content as a data point. After all, "time spent engaging with the content" is the revenue driver.