Advertisers fled YouTube after reports its comments enabled exploitation of minors

On 17 February 2019, video blogger Matt Watson posted a 20-minute clip detailing how YouTube's comments were being used to identify and timestamp videos of young girls in seemingly innocent activities, with the platform's recommendation algorithm then steering viewers to similar videos. By 20 February the video had been viewed more than 1.7 million times, and advertisers whose ads ran against the videos began pulling out of the platform.

Nestlé paused all its US companies' YouTube advertising; Epic Games and Germany's Dr. August Oetker postponed spending after their ads were shown to play before the videos; Walt Disney also withheld its spending, according to people with knowledge of the matter. YouTube said any content or comments endangering minors was abhorrent and prohibited, and that it had deleted accounts and channels, reported illegal activity to authorities and disabled violative comments. Total ad spending on the videos mentioned was under $8,000 in the previous 60 days, with refunds planned.

It was YouTube's second advertiser exodus in two years, after the 2017 pullout over ads appearing next to extremist and violent content; brands including Procter & Gamble and AT&T had since returned. YouTube released an updated policy on 19 February about handling content that 'crosses the line', but the episode showed the same underlying exposure: adjacency to user-generated comment ecosystems it did not fully control.

Disney, Nestlé, Epic Games and Dr. Oetker paused or withheld spend; YouTube planned refunds on under $8,000 of affected ads — its second advertiser exodus in two years.

Moderation treated videos, not comments, as the unit of risk — so a legal video became unsafe through the conversation underneath it.

Recommendation algorithms amplified the pattern, linking exploited clips to each other faster than reviewers could find them.

YouTube had already sold advertisers a fix once after 2017; a second child-safety scandal in two years undercut that reassurance.

The comments section is part of the product: an algorithmic recommendation layer on top of unmoderated comments can turn any innocent upload into a hazard advertisers won't sit beside.

YouTube deleted accounts and channels, disabled violative comments, reported activity to authorities and promised refunds; it released an updated content policy the same week. The material records no quantified revenue loss.