History of the problem

2018 and earlier

World Animal Protection's dataset of the most-viewed staged rescue videos begins in October 2018, suggesting the genre was already established by then. National Geographic later reported that animal welfare groups had first started noticing these videos "several years" before its own 2021 investigation. No single origin video or channel has been definitively identified.

2020

Lady Freethinker conducted a three month investigation into YouTube, identifying 2,053 videos across roughly 150 channels with 30.8 million combined subscribers, depicting animals deliberately harmed for entertainment. The videos had over 1.2 billion combined views. A sample of 6 channels hosting staged rescues specifically had 41.7 million views across 17 falsified videos. Lady Freethinker estimated the content could have earned creators up to 15 million dollars, and YouTube up to 12 million dollars in ad revenue.

The Guardian covered the findings in December 2020, prompting YouTube to remove some flagged videos, though 91 percent of the original set remained live by the end of the investigation.

2021

On March 25, YouTube publicly pledged to ban fake animal rescue videos, telling The Times it would treat them as a "disturbing trend" and begin enforcement within weeks.

On June 30, National Geographic published its own investigation into the genre, featuring wildlife biologist DJ Schubert of the Animal Welfare Institute and photographer Brent Stirton, both of whom described the dramatic camera angles and sheer volume of these videos as tells of staging. YouTube confirmed to National Geographic that it was expanding its violent and graphic content policy that same day to explicitly cover staged rescues.

Days later, World Animal Protection published "Views that Abuse", finding over 180 fake rescue videos uploaded between October 2018 and May 2021, with the 50 most viewed alone totaling more than 133.5 million views. Critically, the report found that between March 26 and June 1, in the roughly nine weeks immediately following YouTube's pledge, 47 new fake rescue videos had already been uploaded across 15 channels, together drawing more than 7 million views.

In August, the Social Media Animal Cruelty Coalition (SMACC), a coalition of animal welfare organizations, published its first joint report, "Making Money from Misery", documenting 5,480 links to animal cruelty content across YouTube, Facebook, and TikTok with a combined 5.3 billion views.

In October, Lady Freethinker sued Google in Santa Clara County Superior Court, alleging breach of contract for failing to enforce its own community guidelines against animal abuse content, including staged rescues.

2022

Meta began working with SMACC in June to help identify and flag primate abuse and fake rescue content. TikTok began a similar collaboration with the coalition in August.

2023

In January, a Santa Clara County judge dismissed Lady Freethinker's lawsuit against Google on Section 230 grounds, ruling the claims ultimately sought to treat Google as the publisher of third party content. Lady Freethinker appealed.

In September, SMACC published a report on pet macaque abuse content, documenting over 1,200 links across Facebook, Instagram, TikTok, and YouTube with more than 12 billion combined views, over 60 percent containing overt physical abuse.

Euronews and other outlets continued to cover the ongoing scale of the problem through the year, noting that the same animals often reappeared across multiple staged videos.

2024

On June 3, the California Court of Appeal affirmed the dismissal of Lady Freethinker's lawsuit, holding that Section 230 barred the claims regardless of how they were framed as contract based.

On October 1, SMACC published "Spot the Scam," its most detailed report on the genre specifically. Over a six week investigation, it documented 1,022 links to fake rescue content across Facebook, Instagram, YouTube, TikTok, and X, with a combined 572,013,959 views. Almost 52 percent of the content was hosted on Meta's platforms. Cats were the most featured animal, at 42 percent of videos. Roughly 21 percent of creators solicited donations directly, and about 22 percent of sampled links had been surfaced to researchers by the platform's own recommendation algorithm rather than manual search.

2025

Lady Freethinker documented more than 2,300 animal cruelty videos across Facebook, Instagram, TikTok, X, and YouTube over a two month window, collectively drawing over 1 billion views. Staged fake rescues were the second most common category found, after monkey torture content. In December, Lady Freethinker, as a Meta shareholder, filed a shareholder proposal calling on the company to address the volume of cruelty content on its platforms and publish its moderation practices.

Separately, AI generated fake rescue videos began appearing as a distinct new problem, including fabricated wildlife rescues with no real animal involved at all.

2026

The problem remains widespread and has shifted further onto Meta's platforms relative to YouTube and TikTok. Enforcement is still inconsistent, and platform recommendation systems continue to surface this content to people who never searched for it.

AI cuts both ways here, and the two ways aren't symmetrical. The obvious risk is that fully synthetic "rescue" videos let creators fake the whole scenario with no real animal involved, and wildlife agencies have begun issuing guides on identifying exactly this kind of content. But there's a more interesting possibility underneath that: if audiences and ad revenue respond to AI generated rescues the same way they respond to staged real ones, creators have no reason left to actually put an animal in danger. The economic incentive that created this entire genre, ad revenue tied to a dramatic thumbnail, could be satisfied without harming anything real. That would be a genuine win, not just a lesser evil.

The catch is that this only plays out as a substitution if the two are kept separate: viewers and platforms need a way to know a video is synthetic rather than staged-but-real, and revenue needs to actually follow the synthetic version. New AI watermarking and disclosure laws could help exactly this problem, giving platforms a reliable signal to route ad revenue and trust toward disclosed synthetic content instead of leaving the distinction to chance.

But watermarking only helps if the audience for this content actually wants what synthetic video offers. Some of the appeal seems to depend on believing an animal was genuinely in danger, not just watching a well made animation of one. If that's true, a labeled AI video may simply draw fewer views than an unlabeled staged-real one, in which case the incentive to actually endanger an animal never goes away, regardless of what the law requires creators to disclose. Whether AI reduces real animal harm or just adds a second stream of deceptive content likely depends on how large that gap in appetite turns out to be.