Japan town retracts latest AI bear image
The animals have killed a record 13 people this year
[TOKYO] A Japanese town deleted a social media post warning of a bear sighting after discovering that a picture it had received showing the fearsome creature was artificial intelligence (AI)-generated.
Similar fake images have been circulating online as fear of bears runs high in the country, where the animals have killed a record 13 people this year.
“The town prioritised informing residents to avoid danger, but we apologise for causing any anxiety or confusion,” the northern town of Onagawa said on its official X account on Wednesday (Nov 26).
The image created with AI showed a bear roaming around a residential area at night.
“We will take this experience as a lesson, and will strive to improve the accuracy and speed of our future information dissemination,” the town said.
But residents still “need to continue exercising utmost caution regarding bear sightings”.
An official in Onagawa said on Thursday that the town had received the bear picture from a well-meaning company president on Wednesday morning.
“There had been reports of a bear sighting in a different district of the town over the weekend, that we warned of over disaster prevention radio, so we didn’t have much doubt” about the image, he said.
The town posted the image on X quickly on Wednesday morning, because the alleged sighting was near a nursery school, the official said.
Schoolchildren in the town were told to commute in a group or use school buses, while nursery school preschool children refrained from playing outside.
At the same time, “we were checking with different apps if the image was genuine or fake”, said the official, who requested anonymity.
“One analysis showed the possibility was high that it was AI-generated, while the other said the possibility of AI was low,” he said, highlighting the difficulty of spotting increasingly realistic-looking AI images.
Image created for fun
It emerged that the image was originally created by a company employee for fun, but one of his colleagues believed it was real and reported it to the boss.
The town retracted its post with the image on Wednesday afternoon, after being contacted by the person who created the AI image, the official said.
It is not the only AI-generated image that has gained traction in Japan as anxiety grows over bear attacks.
There has been a steady flow of genuine reports of bears entering homes, roaming near schools and rampaging in supermarkets, especially in rural northern regions.
When reporters at the Yomiuri Shimbun national daily searched for the words “bear” and “video” on TikTok, they found that around 60 per cent of 100 clips analysed were fake.
Some of them had been produced using OpenAI’s video generation tool Sora, the newspaper said this month.
The fake videos included one in which an old woman fed apples to a bear, and another in which an unarmed high school student fended off a bear with her bare hands.
Another showed a bear making off with a dog in its jaws.
Some of them had been watched hundreds of thousands of times, the report said. AFP
Decoding Asia newsletter: your guide to navigating Asia in a new global order. Sign up here to get Decoding Asia newsletter. Delivered to your inbox. Free.
Share with us your feedback on BT's products and services
