On March 26, 2026, WEBTOON Entertainment announced the first official AI translation program for manhwa. Seven languages, launching spring 2026. Arabic isn't one of them. The real story isn't that AI is coming to comics — it's that it has already arrived, and quietly chose who it serves and who it skips. This piece is an attempt to make sense of that, with numbers instead of slogans.
What WEBTOON actually announced in March 2026
The press release came from WEBTOON's investor relations team — the Korean-American company listed on Nasdaq, with 2 billion downloads worldwide — and carried a deliberately boring title: "Unified International CANVAS Platform." Underneath the title was a quiet bomb — a machine translation program bolted onto CANVAS, the arm where independent creators upload their work. The rollout begins spring 2026 as a beta, and only for English-speaking creators.
The seven launch languages: English, Spanish, French, Indonesian, Thai, Traditional Chinese, German. Seven. No Arabic, no Turkish, no Russian, no Hindi. A company with 2 billion downloads picked seven and called it done.
WEBTOON pressed two defensive points: the program is fully opt-in, and creators' visual artwork will not be used to train the model — only text gets translated. The reassurance makes sense after two years of AI battles with artists. But the timing isn't a coincidence.
Japan didn't wait for WEBTOON: Orange, Mantra, and the $19.5 million war
WEBTOON is late. That part needs underscoring.
In May 2024 — nearly two years before WEBTOON's announcement — Japan's Orange Inc. raised 2.92 billion yen, roughly $19.5 million, in a Pre-Series A round. The round wasn't led by some obscure tech outfit; it was led by Shogakukan — one of the largest manga publishers in Japan. The stated goal, and it's a number worth pausing on: 500 manga volumes translated per month. That's five times the industry's traditional rate. 50,000 volumes over five years is the full ambition.
Then came Mantra Inc. 780 million yen in a single round, from Shueisha, Shogakukan, KADOKAWA, and Square Enix. Four of the biggest names in Japanese publishing put their signatures on one check. This isn't speculative investment — it's strategic technology acquisition.
The Mantra engine doesn't sit in a lab. It's the one translating ONE PIECE and SPY×FAMILY into Vietnamese on MANGA Plus by SHUEISHA, which now supports nine languages and processes around 200,000 pages per month through the Mantra engine. Machine translation is no longer at the margins of the industry — it's inside the chapters that millions of readers consume every day.
The clear read: when WEBTOON announced its program in March 2026, Japan had beaten it by two years with checks rather than press releases. The big publishers settled the debate internally and paid up. The question is no longer whether AI will enter the industry, but who profits from its arrival — a debate that echoes Crunchyroll's slowdown against the automation wave hitting the broader industry.
Artists and translators are saying something different
From the other side of the checkbook, the picture looks completely different.
In June 2024, the Japan Association of Translators (JAT) issued a formal statement calling machine translation of manga "wholly unsuitable for highly contextual texts such as manga," warning that rushing automation "will weaken Japan's soft power." It wasn't an emotional outburst — it was a union position signed by a professional body decades old.
Two months earlier, in April 2024, a new organization was founded under the name Freelance League of Japan (FLJ), as a direct response to the AI wave. Their stated reasons: moral rights of authors, copyright, and "the threat of AI" — written exactly like that, in their founding charter.
Then come the Korean numbers, and they're more painful to the optimistic narrative. A 2024 survey by KOCCA (Korea Creative Content Agency) covered 800 webtoon artists. The results:
- Only 18.3% of artists had tried generative AI.
- Only 36.1% planned to use it going forward.
- Against that: 63.8% of webtoon companies planned to use it.
The gap between 36.1% and 63.8% isn't a technical statistic — it's the full fault line running through the industry. Companies are paying; creators are pushing back.
Why are artists resisting? 41.3% said a lack of expertise in AI copyright and ethics deterred them from using it. 31.3% feared losing the originality of their work. These aren't abstract anxieties — they're numbers pulled straight from the mouths of the people drawing the chapters we read.
When Viz Media failed: the Mikadono Sisters lesson
There's a moment from 2024 we should remember every time someone says AI is "ready."
Viz Media, the U.S. arm of Shueisha, published an English translation of "Dealing with Mikadono Sisters Is a Breeze" using Orange's engine instead of a human translator. The reader response was swift and disastrous: a flood of negative reviews, social media campaigns, and direct accusations that the text "loses the humor and distorts the characters."
What does the incident prove? Not that the technology is bad. That it succeeds at conveying meaning and fails at conveying the breath — the character's voice, a passing joke, a cultural reference that needs 12 seconds of a human translator's thought to land on the right equivalent. And that's precisely what gets a reader to come back every Thursday to the same series.
Orange itself absorbed the lesson. In September 2024, when it launched its e-commerce store emaqi in North America in partnership with Shueisha, it announced with striking candor that "less than 10% of content at launch will be translated with AI assistance." An implicit admission that reader trust is fragile — and that anyone pushing automation too fast loses a market before they win it.
Arabic is off the map — again
Go back for a moment to the seven languages WEBTOON picked: English, Spanish, French, Indonesian, Thai, Traditional Chinese, German. They share one common denominator that doesn't need elaborate analysis — mature digital payment markets, published copyright frameworks, and legal infrastructure that understands Stripe subscriptions and advertising contracts. Arabic isn't like that in most Arab countries. Not because Arabic readers count for less, but because the digital infrastructure is still taking shape.
Industry reports name the Arabic and Turkish markets outright as "underserved" markets dominated by unofficial translators via scanlation. 400 million Arabic speakers, zero official announcement. The logical next market, but not the first.
What does this mean for Arabic translation teams on aggregators? Two opposing pressures. On one hand, no immediate official competition threatens their work; WEBTOON won't be machine-translating into Arabic in 2026. On the other hand, when official AI does reach Arabic — and it will, because the global webtoon market is estimated at $10 billion to $14 billion in 2026 and projected to hit $60.25 billion by 2031 — it will arrive fast, with full legal backing and well-positioned distribution partners. There won't be much time to adapt.
The question worth asking — the one that decides how we read in 2028 — isn't "will AI arrive?" but: when it does, will we have official Arabic platforms paying human translators as a quality-assurance layer, or will we leap straight from chaotic scanlation to machine translation with no human in the loop? For anyone who wants to gauge the size of the Arabic readership awaiting this transition, our guide to the top 5 manhwa of 2026 gives a picture of the series Arabic readers are actually reading today.
So: will AI kill the manga and manhwa industry?
The short answer: no. The longer answer deserves two paragraphs.
What won't happen: drawing won't die, stories won't disappear, and big creators won't be replaced. The KOCCA numbers are clear — 81.7% of Korean artists haven't touched AI at all, and the industry is growing at close to 20% annually with or without AI. The idea that an algorithm will write the next chapter of One Piece is media entertainment, not industry analysis.
What will actually happen: the middle-tier translator's role will get squeezed in the first category of titles. Human translation will stay for the major works (One Piece, Solo Leveling, Jujutsu Kaisen) and for dense-context series. The daily chapters of secondary titles will gradually shift to the machine track with light human review. And this isn't the start of the story — Naver Webtoon has been running its Webtoon AI Painter tool for automatic coloring since 2021, and is developing AI Editor and Auto Drawing. Automation didn't begin with translation, and it won't end there.
For us as Arabic readers, this means three practical things: track what's happening in the served markets because it's coming to us, pressure the official platforms to include our language before they pick 14 others ahead of us, and demand — when AI does arrive officially — transparent human review on top of it. Machine translation without human review means the Mikadono experiment, and we've seen how that ended. Meanwhile, the MangaTime reader's guide for 2026 remains the recommended path for the reader who wants to stay at the industry's actual core, not its media margins.
The industry's other shifts — platform expansion and webtoon's move into streaming — tell the same story from a different angle, as in the case of Viral Hit lands on Netflix amid a storm of criticism. AI is one layer in a larger transformation, and anyone who wants to follow it regularly can find the rest in all manga and manhwa industry news on MangaTime.