Photo/Illutration A fake video featuring Prime Minister Fumio Kishida’s image and voice (Captured from a post on X)

A fake video of Prime Minister Fumio Kishida making vulgar statements in a news program went viral on social media before its creator deleted the footage, which he said he made for fun.

Nippon Television Network Corp. (NTV), whose program logo was used, was anything but amused and alerted viewers about the content.

The video, which was apparently created using generative artificial intelligence and purported to show Kishida talking to camera, had the prime minister’s oversized face superimposed on a person wearing a suit. The footage used the logo of NTV’s “Nippon TV News 24” news program and pretended it was being broadcast live.

“It is unforgivable that the broadcast and the program logo were misused in the fake video,” NTV said in a later program.

The video appeared on social media platforms such as X, formerly known as Twitter.

Its 25-year-old creator told The Asahi Shimbun, “I had no intention of insulting or attacking the prime minister personally but used his image as a celebrity icon.”

The man, who lives in Osaka Prefecture, said he made the video in less than an hour using a free tool.

He said Kishida’s voice was generated by having generative AI learn a video of Kishida speaking.

He said the image was taken from an NTV news program that appeared close to the top of a list of search results with the combination of keywords “Prime Minister Kishida” and “News.”

“I thought I had struck a balance because I showed I was joking by making his mouth open and close in an obviously unnatural manner,” the man said. “I sincerely apologize for using an NTV logo.”

He said he had deleted the video.

The man said he had often created and posted videos featuring politicians, celebrities and anime characters as a hobby.

While he accepted that criticism is the price he must pay for Kishida’s fake video, the man complained that, “What is visibly ‘graffiti in a toilet’ has been cut out in a context that is out of my hands, and I feel I have no control over it.”

G-7 CODE OF CONDUCT

Group of Seven industrialized economies, which aim to create international rules on artificial intelligence by the end of this year under the Hiroshima AI Process, agreed on a code of conduct for developers of generative AI on Oct. 30.

The code of conduct calls for identifying risks of generative AI and implementing appropriate measures. The risks include promotion of false information and invasion of privacy.

(This article was compiled from reports by Asako Myoraku and Shiori Tabuchi.)

Related News