TikTok knowingly lied about how its algorithm led to a suicide 74%
By Hillary Keverenge88%
8/6/2026, 5:38:47 AM
BS Summary: This article contains 25 faulty reasoning types, including Hasty Generalization, Unattributed Quote, and Appeal to Emotion, with Negativity Bias as the most egregious example at 64.8% saturation with 162 hits. Analysis detected 1,175 faulty-reasoning hits from 250 analyzed words, generating a BS Score of 58.1% and a BS Rank of 74% (7,809 of 29,795 articles). This article is worse (more manipulative) than 73.80% of the article peer group.
A leaked internal report reveals TikTok withheld algorithm safety features from ~15 million people in 2022 to test how guardrails impacted daily active user (DAU) metrics.
The report notes that a 16-year-old was randomly placed in the unprotected control group, flooding his feed with thousands of suicide and self-harm videos shortly before his death.
TikTok says user safety remains a top priority, and it continues to invest in Trust and Safety.
TikTok is certainly no stranger to controversy.
Earlier this year, alarming shifts in TikTok’s privacy policy had many of us weighing whether it was finally time to hit delete on our accounts .
Yet, as anyone who has searched for a decent TikTok alternative knows, finding a platform that recreates its hyper-personalized magic is far easier said than done.
Now, a devastating revelation regarding how the app handles user safety might force even the most dedicated scrollers to think twice about their continued presence on the platform.
According to a confidential internal report obtained by Bloomberg , when TikTok updated its algorithm in 2021 to stop users from getting overwhelmed by harmful content, the company intentionally withheld the fix from millions of users.
To measure how safety features might impact app stickiness and daily active user (DAU) metrics, TikTok conducted a massive A/B experiment.
By design, a control group comprising 10% of US users — about 15 million people — kept the old, unprotected version of the algorithm.
Speakers
1speaker6.8%attributed speech233writer words
Selected voice
100%flagged-word coverageTikTok
17 attributed words100% of attributed speech100% writer coverage
Attribution is sentence-level. Pattern percentages are calculated only from words assigned to that voice.
Loading…
Loading…
Loading…
Loading…
Analysis
Hover over highlighted words in the article to view the associated bias or fallacy analysis.