When a studio known for charm and craftsmanship gets accused of “AI slop,” the reaction tends to be loud. That’s exactly what has happened to Level-5, the developer behind Professor Layton and Yokai Watch, and the fallout is still unfolding.
I run Pale Shadow Gaming, where the goal is to let games unfold naturally, or in my own words, “letting the moments happen naturally rather than forcing a script.” Authenticity matters to me, which is partly why this story caught my eye. The details below come from Verity Townsend’s report for IGN, paraphrased in my own words.
What Happened
On September 10, Level-5 published its “VISION 2026 II: Dreams” showcase on YouTube, narrated by a cartoon version of CEO Akihiro Hino. It covered upcoming titles including Yokai Watch 2: Haunted Domain, Decapolice, and a first look at Professor Layton and the New World of Steam.
Almost immediately, viewers began picking the video apart. Fans scrutinized backgrounds and artwork, flagging inconsistencies they believed pointed to generative AI. Some compared the Yokai Watch 2 trailer from the September 9 Nintendo Direct with the version in Level-5’s showcase a day later, claiming to spot AI-like errors. Sleuths also suspected that backgrounds in the Direct trailer for the Professor Layton and the Curious Village remake were AI upscales of the original art. Even a concept image for the studio’s Tokyo Game Show booth got flagged.
The First Response Landed Badly
Hino addressed the criticism on X, admitting the team had experimented with AI processing in the showcase out of a wish to make it flashy. He apologized for the offense, but insisted that core elements like story, character designs, and settings are made by people, and that AI was used to speed up digitizing artists’ work. He also said AI-generated data isn’t present in released games and that efficiency gains give developers more time to be creative, as the studio tries to shrink a roughly five-year development cycle for major titles down to two.
That explanation didn’t calm anyone. The idea of using AI to make a presentation flashier struck many Japanese commenters as style over substance. Others raised ethical objections about how generative AI is trained on artists’ copyrighted work. One commenter argued that even a single AI-made asset leaves fans permanently wondering whether a song, image, or line of dialogue is human, and that preserving trust matters more than release speed.
The Second Statement: Apology Plus Explanations
On September 17, Level-5 issued a further statement apologizing for the distrust caused by what it called its inexperienced use of AI and simple human errors. It repeated that AI hasn’t been used directly in the games’ contents, promised more care about accuracy and quality, and said it would fix both AI-related errors and human mistakes that merely look like AI errors. Corrected trailers are planned to coincide with Tokyo Game Show, and versions of the Decapolice and Holy Horror Mansion trailers had already been uploaded.
The company also went through specific complaints:
- Telephone lines in a Yokai Watch 2 illustration: Level-5 said they were hand-drawn by an in-house designer based on the artist’s own photo, deliberately simplified, and misread as AI. They’ve since been adjusted to look more natural.
- Vending machine and guttering: The company called these human errors, details that were accidentally left out.
- Bicycle brake levers: Previous Yokai Watch games didn’t draw them, but the studio felt they belonged in a game with this much detail.
- Nathan’s changing face: Level-5 said no AI was involved. It couldn’t fix his low-poly model before the Nintendo Direct submission deadline but had updated it by its own showcase the next day.
Level-5 also stated that no AI was used in Professor Layton and the New World of Steam. It didn’t address whether AI was used in the Curious Village remake. Going forward, the studio says it will limit AI to efficiency and avoid using it directly in videos released to fans.
Why the Reaction Is Still Sour
For many, the apology missed the point. One commenter argued the anger was never about small errors; those were merely clues that AI was involved, and the real objection is to using it at all. Their worry was that Level-5 will keep using AI and just get better at hiding it. Japanese commenters raised a different concern, that blaming human error shifts responsibility onto staff, and that the real issue is the president’s repeated promotion of generative AI without addressing problems like unauthorized data use and environmental cost.
It’s worth noting this isn’t new territory for the studio. Level-5 gave a presentation in December 2023 to a Japanese government study group about its AI use. It described using Stable Diffusion to generate layout ideas for promotions and to build 3D versions of 2D models from older games like Yokai Watch, with human artists producing the final output. It also described combining parts of AI-generated images with artist-made work for backgrounds such as stadium crowds and buildings in Inazuma Eleven: Victory Road.
What to Watch For
The key question is how much damage this does to player trust. Reuploaded trailers and promises of restraint may fix individual videos, but they don’t settle the bigger debate. Fans want to know where the line is, and whether future disclosures will be transparent or merely more polished. With Tokyo Game Show underway and Level-5’s next games approaching, the studio’s actions over the coming months will say more than any statement.
article by paleshadowgaming
