For 35 days, Spanish fact-checking outlet Maldita.es investigated networks that use TikTok to trade or sell child sexual abuse imagery. Its reporters say they logged more than 1,000 abuse videos and identified 26,000 accounts potentially tied to those networks. Much of the flagged content was still up when the story ran, according to the investigation.
On the morning of Nov. 5, the reporters also tracked which ads showed up in their feed. Over two hours, they counted 21 ads, including ones for Spanish chocolate drink brand Cola Cao, Adobe, Samsung, Colgate, Milka, Booking and Apple. Each ran immediately before or after one of those posts.
In some cases, the video itself showed abuse. In others, it came from an account selling that kind of imagery. The reporters also saw ads next to videos of minors who were being sexually harassed in the comments.
The reporters were browsing with the account they had used since the start of the investigation, and its history may have shaped TikTok's recommendations. That means the two-hour sample documents specific placements, not how often a brand is likely to land next to this kind of content.
That was enough to get several advertisers to act. Spanish bank Santander and Spanish energy company Moeve confirmed to Maldita.es that they had suspended all of their TikTok campaigns worldwide. Department store chain El Corte Inglés and its insurance arm also paused their campaigns, according to the outlet. Spanish trade publication Reason Why reported that Nestlé paused its Nescafé Dolce Gusto campaigns and that Samsung suspended its advertising in Spain. Moeve, security company Verisure, Spanish retailer Eroski and Starbucks also stopped posting on their own TikTok accounts. Eroski said it wants to review the platform's safety assurances before it returns.
On paper, media buyers have several ways to control where their ads run. Before a campaign goes live, TikTok's Inventory Filter lets them pick one of three inventory tiers. Even the most permissive tier is supposed to keep out explicitly inappropriate content. But TikTok says in its documentation that despite "reasonable efforts," it can't guarantee those exclusions will hold for every impression.
Once the campaign is live, advertisers can bring in DoubleVerify, Integral Ad Science or Zefr to check which videos ran before or after their ads. Maldita.es doesn't say whether the brands involved had turned on a filter or brought in a third-party verification vendor to monitor their campaigns.
What brand safety tools miss beyond the video itself
Sometimes a video shows nothing explicit on screen. The problem is in the comments, where users offer abuse imagery, or in the account's bio, which points to a distribution network. A tool that only analyzes the video can rate it brand-safe and miss what tipped off the reporters.
The investigation also looks at how users find these posts. Maldita.es describes how TikTok's search and recommendation systems steered it to the accounts it reported. For brands that have also pulled their organic posts, coming back isn't just a question of where their ads land.
Advertisers have been here before. In February 2019, Nestlé paused its YouTube campaigns after predators' comments turned up under videos of children. YouTube then disabled comments on tens of millions of videos. The parallel to TikTok is clear: an ad could run on a harmless-looking family video, and only the comments showed what the brand was sitting next to.
In July 2020, Facebook's moderation policies sparked the Stop Hate for Profit boycott. At the urging of U.S. civil rights groups, advertisers including Unilever, Verizon and Honda paused spending on the platform to pressure Facebook to crack down on racist and hateful speech. Organizers say more than 1,200 companies and organizations backed the campaign. Brands that have also stopped posting organically on TikTok are now taking a similar public stand.
How long do budgets stay frozen? By late July 2020, The North Face, one of the first brands to join the Facebook boycott, had already announced it would return in August. Facebook's ad revenue grew 22% year over year in the boycott quarter. YouTube came through its 2019 crisis just fine, too: its ad revenue rose 36% that year, to $15.1 billion. For TikTok, the number to watch is not how many brands announce a pullback this week. It's how much budget is still on hold three months from now.
Is there a double standard between social and the open web?
Say a news site publishes an investigation into child sexual abuse on TikTok tomorrow. Those words alone can get the article blocked from campaigns. That's the brand suitability debate: an advertiser can decide a topic is too sensitive for its campaign, even when the reporting is beyond reproach. In April, IAS made that case using Reuters as an example. In its study, keyword filters blocked 54% of the articles that its contextual analysis rated brand-suitable.
On TikTok, meanwhile, Maldita.es went from a post where minors were being harassed to a major brand's ad in a single swipe. That goes to the heart of brand safety: what advertiser would sign off on running next to that? Here's the paradox for publishers: the article that exposes the abuse can lose an ad impression because of the words it uses, while the feed where the abuse happens sold the impression.
Spanish prosecutors take TikTok to court
On Nov. 13, the case moved to the courts: Spanish prosecutors filed a complaint against TikTok Technologies Spain in a Madrid court. They asked the court to investigate, determine any liability and order the reported content taken down immediately.
Spain's minister for youth and children, Sira Rego, called for TikTok to be suspended in the country as a precaution. Spain's data protection authority announced narrower steps that target specific content and links.
TikTok says Maldita.es didn't share all of its findings and argues that this makes the content harder to find and take down. The outlet says it used TikTok's reporting tools and handed its evidence over to authorities.
What about France?
Santander and Moeve didn't give a country-by-country breakdown of the budgets affected by their global pullback.
As of Nov. 17, no France-specific suspension had been announced in response to the investigation, according to the public sources we reviewed. But France does come up in Maldita.es's follow-up checks on its initial findings.
On Nov. 10, the reporters found that some searches already blocked in Spain were still returning results in France. Media teams that buy across several markets will need to close that gap before campaigns restart.
What can TikTok show them, country by country, about what it has done since the reports came out? For brands whose ads appeared in the investigation, the answer starts with the impressions Maldita.es documented. Knowing which controls were in place when the ads ran, and what they would catch today, would give them something concrete to weigh before deciding whether to return.


