ByteDance Makes Use Of A Moderation System That Entails Word Lists To Detect And Suppress Sensitive Content

TikTok’s parent firm ByteDance is making news again after it utilized a moderation system comprising word lists for the detection and suppression of certain content.

And the content in question is dramatic including matters linked to China, Donald Trump and his fiasco during the American elections, Uyghurs, and more. How’s that for some shocking behavior?

There are also some reports that highlight how the system is designed to keep an eye on YouTube which is TikTok’s rival. But despite all these facts coming out, TikTok has denied that its systems were ever moderated or affected by such lists coming out of its parent company.

Is the news shocking, well, it’s not the only leading social media company that is maintaining a library of this sort. This is used to dictate what users of the most popular platform could and couldn’t see. But this tech giant from China has a huge glossary that goes well beyond the usuals you witness in content moderation blocking lists. Those are designed to focus on hate speech and remove such terms or phrases linked to terrorism or kids’ safety.

This one, on the other hand, is designed to focus more on China, its relations with the US, its trade affairs, former US commander-in-chief Donald Trump, and even how it persecutes a minority ethnic group called Uyghurs. This also entails a few of the social media app’s US competitors that are subjected to the likes of monitoring. And in a few situations, it’s linked to ByteDance suppressing them altogether.

Remember, this kind of monitoring or suppression list is like a tool that entails a range of words it deems specific like Trump-directed prohibited terms, sensitive persons on TikTok, Japanese comments or banned words on TikTok, audio sensitive works in Tibet by TikTok, and the lists go on.

One internal guide for this kind of tool might appear to be written down by Chinese employees from ByteDance and also from its subsidiary in the Chinese capital of Beijing. It ends up describing the huge global core vocabulary and also has some sensitive terminology posing a high risk. It also has its own illegal vocab list in the form of a thesaurus.

The system is not something new and actually came about in the year 2017 when it was designed to witness, recall, and evaluate such terms widely displayed across goods by ByteDance.

This might be linked to it bringing more risks to the firm’s safety, revenue, and reputation.

Leading media firm Forbes mentioned how there were currently 50 different lists in the tool that entailed words such as TikTok across the title and others like internal material that showcase tools getting accessed by the likes of workers from the company in the latter part of 2022.

Meanwhile, a number of other lists put out explicit references to the app’s Chinese rival Douyin, and products hailing from ByteDance itself.


Read next: Israel Is Making Use Of New Facial Recognition Technology For Surveillance Of Palestinians, Warns Amnesty International
Previous Post Next Post