Bald eagle 'massaging' its mate? AI deepfakes collide with the laws of the wild

7 min Click here to listen to this article Share via Close extra sharing options Email Facebook X LinkedIn Threads Reddit WhatsApp Copy Link URL Copied! Print 0:00 0:00 1x This is read by an automated voice. Please report any issues or inconsistencies here. Shadow gingerly places one taloned foot, then the other, on Jackie as she hunkers down on the nest. With Big Bear Lake glittering in the distance, he raises each foot in a kneading motion — evoking a bald eagle massage. “Somehow, it says everything about their bond,” reads the caption on the 15-second video posted to Facebook. Jackie and Shadow — made world-famous by a 24-hour livestream — aren’t the only animals falsely depicted in deepfakes. AI wildlife videos have flooded social media platforms like YouTube, Instagram and TikTok, racking up millions of views and likes. Some are whimsical, like a handful of bunnies hopping on a trampoline. Others take a more menacing tone, like a jaguar facing off with a dog in a snowy backyard. Far from benign, some experts say the videos can skew how people view and even interact with wildlife — potentially leading to perilous encounters. They may also undermine viewers’ growing desire to tune into nature to escape the frenetic rhythms of daily life. Repeated exposure could erode trust in media and institutions generally, with one Reddit user proclaiming, “Can’t even watch real animal videos because 90% of them are AI.” There are also legal implications. The deception works because the depictions are often hyperrealistic. Even a producer for the Dodo, an animal-centric media outlet, admitted to falling for the bouncing bunnies. Often the videos appear to be ripped from trail or security cameras, enhancing vibes of authenticity. In the competitive economy for people’s attention, the videos can help win looks and likes, potentially driving ad revenue for those who post them. Megan Brief, a digital marketing coordinator for Natural Habitat Adventures, an ecotourism company, had just returned from Svalbard, a far-flung Norwegian archipelago teeming with polar bears and walruses. Her social media feed piled up with video after video of polar bear rescues, such as fishermen or scientists hauling a freezing, struggling baby polar bear onto a ship. On board, people snapped selfies with the cub before reuniting it with its mom. She knew they were fake because she was well-versed in the behavior of the snow-white predators, which are fiercely protective of cubs. As the U.S. Fish and Wildlife Service warns, these “large, powerful carnivores” can easily injure or kill people. It would also be illegal to intervene. But thousands of commenters took what they saw at face value. (Photo illustration by Jim Cooke / Los Angeles Times; Source photo / Getty Images) “It shows that you can have this close proximity with wildlife that is not only dangerous to you, but it’s dangerous to the animal,” said Brief, who is also a wildlife photographer. Social media is filled with AI animal rescues of all types. “That’s everyone’s dream, to be one with all the animals and with wildlife,” she added, “but you have to respect their habitat and their behavior and give them the space that they need.” On the flip side, she said the videos also can perpetuate myths that predators such as wolves and mountain lions are more dangerous than they actually are. It’s easy to see how videos could inflame heated debates over managing such animals, in California and beyond. In a paper published last September in “Conservation Biology,” researchers said the videos also can make people think animals are more abundant, or less threatened, than they are. They might donate or volunteer less as a result. “If the public is unable to distinguish between actual threats to biodiversity and fictionalized narratives, the perceived urgency to act may diminish,” the researchers wrote. Jenny Voisard, media and website manager for Friends of Big Bear Valley, a nonprofit that operates cameras trained on Jackie and Shadow, said her inbox is overloaded with complaints about AI content. Grifters are nothing new — the nonprofit has long contended with fake accounts — but they’ve evolved with the technology. People who follow the beloved eagles are fed more content about them by the algorithm, and she said AI rises to the top of the feed. (That seems to explain why this reporter is often served the fakes when opening Facebook.) “People get very upset when they see someone depicting Jackie and Shadow in an unnatural way or wrong, or when it looks like they could be in danger,” said Voisard. Some clips showed owls and ravens attacking the couple, especially riling up fans. The nonprofit recently trademarked its name and is in the process of copyrighting its livestream. She said the point is to protect what they create, such as merchandise and a detailed log of what the eagles are up to, from fakers. However, ownership in the age of AI is fraught. Voisard said their li

Bald eagle 'massaging' its mate? AI deepfakes collide with the laws of the wild

Key takeaways

More in local