The dangers of AI animal “slop” videos | Explained Premium

While an AI-generated video of a man taking an underwater selfie with a shark is easy to identify as fake, more realistic Gen AI videos can warp the way we think about wild animals and their behavioural traits

The story so far: Open up Instagram, Meta AI, TikTok, YouTube Shorts, or X on your phone and it won’t be long before you scroll to a short video clip that shows animals and humans sharing a bizarre encounter.

While an AI-generated video of a man taking an underwater selfie with a shark is easy to identify as fake, more realistic Gen AI videos can warp the way we think about wild animals and their behavioural traits, putting both us and them in danger.

This is thanks to a barrage of Generative AI tools that allow users to enter text prompts in everyday language and churn out hyper-realistic videos at scale that can be easily uploaded online. As platforms like Meta pivot to short-form videos in an attempt to surpass TikTok, while X rewards virality and engagement, users are incentivised to make and share more low-cost, low-effort content.

Many AI videos show unrealistic and violent scenarios that could impact the way children perceive and treat animals | Photo Credit: Instagram

As a result, you might have seen gorillas breaking out of zoo enclosures, crocodiles gobbling up their handlers, dying lions surrendering their cubs to humans, fishermen rescuing polar bears, and monkeys dancing with children.

These videos are generally called “AI slop” due to their low-grade entertainment value and short length. Often shared with sensationalised captions or presented as though they are real news events, viewers inevitably waste their time trying to figure out if the content is real or AI generated.