In a short video widely shared online, Pokemon frolic through a lush green field while OpenAI CEO Sam Altman watches from the sidelines. He then turns to the camera and says, “I hope Nintendo doesn’t sue us.”

The fantastical scene is one of a multitude of artificial-intelligence-generated videos created by OpenAI’s short-form video app Sora 2, and the video speaks to the legally fraught territory that the company is entering.

Named for the Japanese word for “sky” due to the product’s “limitless potential,” according to company lore, the platform was released to a handful of users last week and was quickly flooded with videos featuring iconic Japanese intellectual property (IP), including Pokemon, "One Piece" and "Dragon Ball Z." Such videos, which are only possible to generate because of OpenAI “training” Sora 2 on the work of human creators, have been widely branded “AI slop” by critics.

And Sora 2 is particularly hungry for Japanese intellectual property.

“We'd like to acknowledge the remarkable creative output of Japan — we are struck by how deep the connection between users and Japanese content is!” Altman wrote in a recent blog post.

But legally, this places the company on shaky ground.

Politician and lawyer Akihisa Shiozaki, who quickly drew attention to the prevalence of Japanese IP visible across the platform in a series of posts on X, called the widespread depiction of Japanese anime characters across Sora 2 a “serious issue under copyright law.”

Shiozaki, a Lower House lawmaker of the ruling Liberal Democratic Party, echoed the alarm from others online, whom he characterized as claiming that "AI is devouring Japanese culture."

“The efforts and sensibilities of Japanese creators, who have led the world, are at risk of being disregarded. Beyond shock, I felt a chilling sensation down my spine and deep indignation,” Shiozaki wrote, adding that he had called an urgent online meeting with members of relevant government ministries.

Altman has since responded to the accusations of copyright infringement, vowing to give “more granular control over generation of characters, similar to the opt-in model for likeness but with additional controls.”

Video game maker Nintendo has not named Sora 2 or Open AI publicly but has addressed the backlash.

“Whether generative AI is involved or not, we will continue to take necessary actions against infringement of our intellectual property rights,” the company said in a statement.

Sakon Kuramoto, managing attorney at Kuramoto International Law Office, said that generating and publicly transmitting images substantially similar to existing characters without authorization could constitute a type of copyright infringement.

Cultural Affairs Agency panels have also noted that AI platforms could be held accountable in cases of their product infringing content at a high frequency. OpenAI could potentially be breaching guidance if the company was aware that its AI was likely to generate copyrighted works and did not seek to impose restrictions.

Japanese copyright law differentiates between utilizing copyrighted material for training models and generating AI images or profiting from their distribution.

But Kuramoto said any potential legal breaches by OpenAI would be unlikely to escalate into a legal challenge.

“Considering the actual costs and duration of litigation against OpenAI, it is expected that many rights holders will conclude opting out is faster than filing a lawsuit,” he said.

At present, Sora 2 remains invite-only and requires a code to access, but it has already rocketed to the top of the Apple App Store chart. The app is only currently available in the U.S. and Canada but is expected to be rolled out to other jurisdictions soon.

In March, OpenAI’s ChatGPT faced claims of copyright infringement after a viral trend saw users generate images echoing the style of Studio Ghibli. The animation studio has not legally retaliated.

But copyright is just one concern that has been raised following the release of Sora 2, with experts warning of the risks posed by realistic AI-generated videos that are increasingly difficult to distinguish from reality.