TikTok will let customers tone down the quantity of AI content material of their feed

TikTok is the most recent social platform handy customers extra management over the quantity of AI-generated content material they see. The app is experimenting with a brand new setting that may enable customers to request to “see much less” AI of their “For You” feed.

The brand new toggle shall be obtainable in “the approaching weeks” within the app’s “manage topics” part that lets individuals personalize the forms of movies extra more likely to seem of their feed. In a weblog put up, TikTok says the management is supposed to assist customers “dial issues down” in the event that they wish to see much less AI content material of their suggestions. The replace follows an identical transfer from Pinterest, which has confronted complaints about AI slop drowning out genuine photographs on the platform.

TikTok, like most different platforms, has seen an inflow of AI-generated content material in recent times. The corporate, which requires customers to label such creations, says that there are greater than 1.3 billion movies with these labels. On the identical time, the corporate notes that its present strategies for detecting AI content material aren’t excellent, At the moment, TikTok depends on a widely-used watermarking system referred to as Content Credentials that provides metadata to AI-generated content material. However these indicators can get tougher to detect when photographs or movies are edited in different apps or copied and reshared to different websites, as typically occurs with viral content material.

To handle this, TikTok says that it’s going to additionally experiment with an extra “invisible watermarking” system that would assist it establish and label AI content material extra reliably. “‘Invisible watermarks’ add one other layer of safeguards with a sturdy technological ‘watermark’ that solely we will learn, making it tougher for others to take away,” the corporate defined.