A study reveals that TikTok's feeds display AI-generated content three times more often than those on YouTube.

      Content aimed at kids has the highest rate of AI-generated material on TikTok.

      If you've ever sensed that your TikTok feed contains a lot of inauthentic content, you're not wrong. A recent report from Kapwing reveals that 59% of videos shown to a newly created TikTok account are AI-generated, which is about three times the rate found in a similar analysis on YouTube.

      How does TikTok's issue with AI-generated content compare to YouTube?

      Kapwing created new accounts on both platforms and manually reviewed the first 500 videos presented to each. On TikTok, 294 of those videos were identified as AI-generated, whereas on YouTube, only 104 out of the initial 500 Shorts were deemed AI-generated, resulting in a slop rate of 21%.

      The magnitude of this issue is notable, especially considering that TikTok had marked 1.3 billion videos as AI-generated by November. Kapwing also examined over 10,000 TikTok videos across 20 different categories to understand where AI-generated content is most prevalent.

      Which TikTok categories are overcrowded with AI-generated content?

      Kids’ content led all categories, with 57% of the 2,000 videos reviewed being AI-generated. The tag #cartoonkids was particularly bad, with 97 out of 100 videos being artificial.

      Following closely were the categories of Science and Education, Health, and History, each with AI-generated content rates between 33% and 35%, as these often use animation or voiceover narration rather than real-life demonstrations.

      Conversely, categories like Fashion, Music, and Fitness were relatively unaffected, each reporting rates below 2%, likely because these genres rely heavily on actual on-camera presence.

      Despite TikTok having introduced tools for users to reduce AI content in their feeds, this study indicates that the default content still heavily favors AI. Currently, the responsibility for distinguishing between AI-generated and genuine content largely lies with the viewer.

      Manisha Priyadarshini is a tech and entertainment writer boasting over nine years of editorial experience.

      Threads is enhancing community engagement while allowing users to filter out unwanted content.

      Threads has exited beta with new custom icons and a dedicated hub, along with a new feature called Your Algo that enables users to set topic preferences for a tailored feed.

      Threads is implementing several improvements to its Communities feature and launching a new tool for feed personalization, coinciding with Meta celebrating the platform’s 500 million monthly active users.

      Facebook now offers an answering assistant for all your pressing inquiries, akin to Google Search.

      Meta has introduced a range of AI-driven features, with the highlight being AI Mode, a novel method for obtaining answers to questions within the app using real posts from Facebook Groups and Reels instead of standard search results.

      Another study reveals that a concerning number of children are exposed to harmful content on social media.

      Despite the UK's Online Safety Act being enacted last year, a new study indicates that harmful content continues to reach a significant portion of teenagers and nearly half of all girls on social media each week.

      A year post the enactment of the UK's Online Safety Act, research conducted by the Molly Rose Foundation (as reported by The Guardian) shows that a third of all teenagers in the UK and almost half of all girls encountered content related to suicide, self-harm, depression, or eating disorders on social media within just one week.