Japan’s manga artists can’t ignore AI any more

Imagine if the late Osamu Tezuka, Japan’s revered “Godfather of Manga” for his pioneering artistic techniques and visionary behind the iconic worlds of Astro Boy and Black Jack, were alive today. What would his stance be on generative artificial intelligence (AI)?

His eldest son Macoto Tezka believes that he would have embraced it.

“If my father were alive today, I believe he would have used AI,” Mr Tezka said in 2023, announcing a special project that harnessed generative AI to breathe new life into his father’s distinctive style for Black Jack’s 50th anniversary. 

He added that his father, who died at 60 of stomach cancer in 1989, had “mass-produced manga with very high quality without enough people. With AI, he could have produced even more”.

Japan’s creative sector has been wrestling with the question of the role of generative AI in the production process. Some purists have recoiled at the idea, calling it an affront to hand-drawn artistry. 

But there is also growing recognition that the technology, wielded with strategic guard rails and robust copyright protections, could offer a lifeline to an industry caught between chronic manpower shortages and an insatiable global demand.

For decades, the glittery world of anime and manga has been propped up by cheap labour who toil for long hours at minimum wage – sometimes under the guise of “apprenticeships” – a labour model born in a different era where passion may trump pragmatism.

Yet industry demand has only surged, driven by on-demand streaming platforms like Netflix, which has reported that one in every two users watches anime, the crown jewel of Japan’s creative industry.

Japan’s creative exports hit a record 5.8 trillion yen (S$50.4 billion) in 2023, up from 1 trillion yen in 2010. This makes it Japan’s second-largest export industry, behind only automobiles, with the country having set an ambitious target to grow the export market value to 20 trillion yen by 2033.

Democratising creativity

But here lies the paradox: As generative AI lowers the barrier to creating manga and anime-style content, could it also erode Japan’s hard-won dominance?

Japan’s Ministry of Economy, Trade and Industry raised the alarm in a strategy paper released on June 24.

“With the advent of generative AI, there is a risk that in the medium to long term, Japan’s comparative advantage in fields such as anime and manga will decline,” it said. At the same time, it recognised the need for fundamental changes to embrace digital technologies in creative production.

Generative AI is not a distant spectre but a present reality. Mr Tezka said at a news conference, in response to a question from The Straits Times, that it was a tool that could effectively be wielded to grow Japan’s creative sector.

“The demand for new works cannot be quenched, and AI’s strengths lie in its ability to create in a short time,” he said. “But AI cannot depict the underlying emotional depth that propels Japanese works, and so the human hand is still necessary.”

Yet the slow adoption of AI in Japan reflects deep-rooted scepticism. Individual usage of generative AI in Japan stood at a paltry 26.7 per cent as at March 2025, according to the internal affairs ministry’s annual White Paper released on July 8 – far behind the United States’ 68.8 per cent and China’s 81.2 per cent.

Even among businesses, the gap is stark: over 90 per cent of American and Chinese companies reported using generative AI, with Japan at 55.2 per cent.

Still, pockets of high-profile adoption have emerged. Renowned architect Kengo Kuma’s studio began using AI in 2023 to render 3D perspective drawings from 2D hand-drawn ones. Work that used to take one week can now be done in 10 minutes.

Novelist Rie Qudan caused a stir when her novel Sympathy Tower Tokyo – partially written with generative AI – was given the prestigious Akutagawa Prize in 2024. The book, whose English translation is due for an August release, featured characters who tap AI for advice. Qudan used ChatGPT to derive answers to these questions, faithfully reproducing the responses and using them to drive the narrative forward. One judge lauded the work as “flawless”.

There are, however, concerns that easy-to-use AI tools might result in a flood of formulaic content. And while Japan’s latest start-up unicorn, Sakana AI, has developed new AI programmes including one that renders images into ukiyo-e – an art form that thrived in the Edo era – many AI tools are foreign.

There are also fears that an overdependence on AI could chip away at creative standards.

Is imitation the best form of flattery?

Beyond literature, AI is increasingly used in Japanese design, gaming and fashion. But nowhere is its presence more contested than in the anime and manga industry.

Japan’s comparatively lax copyright laws effectively grant free rein for copyrighted works to be used without permission for the deep learning of generative AI – so long as financial interests of copyright holders are not demonstrably harmed.

This laissez-faire policing has led to AI websites being flooded with works that bear an uncanny resemblance to well-known characters. Voice artistes have, meanwhile, also complained that their voices have been copied. Given that Japan already suffers an estimated 2 trillion yen in damages annually due to global piracy, this clearly poses a further threat.

The ethical tightrope is exemplified by the controversy earlier in March 2025 when ChatGPT-generated images that mimic the je ne sais quoi of Japan’s Studio Ghibli proliferated online.

That same month, ChatGPT operator OpenAI also released a feature allowing users to generate four-panel comic strips with character dialogue from detailed prompts.

OpenAI says it “avoids imitating the style of individual living artists” but “permits broader studio styles” – in essence saying that Studio Ghibli as an entity was fair game but not that of co-founder Hayao Miyazaki, 84.

But Studio Ghibli’s distinctive artistry style – epitomised by delicate and emotional storytelling that has defined modern Japanese animation – is inseparable from Miyazaki and his late fellow co-founder Isao Takahata. The latter has an ongoing career retrospective exhibition in Tokyo, which shows hand-drawn sketches from early on in his career to his eventual adoption of digital colouring methods that were unprecedented for its time.

The legendary studio behind cinematic masterpieces such as Spirited Away (2001) and Howl’s Moving Castle (2004) has refused to comment on the fad, which has since fizzled out. 

But the controversy has sharpened awareness of how easily AI can imitate signature styles – without capturing their soul.

Miyazaki once famously said in 2016 that artificially generated images were “an insult to life itself”. His son Goro, meanwhile, reportedly said that while generative AI was “inevitable” in anime, creators like his father were “irreplaceable”. He warned that unless the industry evolves, it will struggle to attract the next generation of talent.

Even if there is a cultural divide between AI-friendly creators and staunch traditionalists, it is ultimately a decision that individual creators will make on their own, just as it is a professional preference to render images by hand or by computer. If Japan were to entirely shut out AI, however, it would certainly lose pace with other film industries worldwide.

Working with AI

That dilemma is echoed by Shirow Masamune, 63, who predicted a futuristic society driven by AI in his cyberpunk series Ghost In The Shell released in 1989.

The reclusive Masamune, who has an ongoing exhibition in Tokyo to mark 40 years since his debut, neither owns a smartphone nor maintains a social media presence.

He told the Yomiuri newspaper in a rare interview: “The convenience of AI and its possibility of causing harm are two sides of the same coin. While it can open up the possibility of a brighter future, it can also be abused or used to incite people more easily than before.”

This underscores the need, he said, to “develop AI correctly”.

Japan’s creative industries have a history of initial unease with innovation – whether it was novelists using typewriters, architects swopping pencils for software, or manga artists colouring digitally. Generative AI is simply the latest disruptor.

Rather than fret about the threat, Japan should focus on what sets its creators apart, including the ability to create original, deeply emotional and culturally rich stories.

“My personal opinion is that both AI and humans are the same in how they study and learn from the masterpieces of those who walked the path before,” Mr Tezka said, adding that while AI may expedite the process, people have long tried to learn and imitate.

The onus is not just to protect, but to adapt. That means investing in ways to fight copyright infringements, such as by tapping the potential of AI and working with overseas jurisdictions in what has long been a cat-and-mouse chase.

Japan could also invest in developing AI tools for anime and manga which are governed by domestic laws and shaped by its cultural values.

It would be futile and self-defeating to go against the grain of generative AI, especially given the pressures that the anime and manga industries face. 

When used wisely, AI can free up creators from routine tasks and help them do what they do best: scale the heights of imagination and tell stories that leave their mark on the world.