Level-5 has apologized for the generative AI controversy arising from its recent online showcase and announced that it will avoid directly using AI in videos released to the public in the future.
On September 17, Level-5 announced measures regarding generative AI and plans to revise several promotional videos shown at its recent showcase via its official website. The company stated that some PVs presented during the showcase contained errors from AI usage as well as human errors that could be mistaken for AI, adding that revised versions of the videos will be re-released in time for the opening of Tokyo Game Show. Furthermore, the studio explicitly clarified that "there is absolutely no direct use of AI" within the games themselves.
The controversy erupted following "LEVEL5 VISION 2026 II" held on September 10. Viewers pointed out scenes in some of the revealed videos that appeared to use generative AI, leading Level-5 CEO Akihiro Hino to admit on his official X account on September 12 that AI was indeed used in the presentation footage. Hino apologized, explaining that experimental processing with the latest AI was applied to make the showcase more spectacular, acknowledging that it was undeniable AI had been used in various ways during the presentation.
However, he explained that elements directly tied to the charm and identity of released titles—such as scenarios, character designs, and fundamental settings—are all created entirely by humans. He emphasized that the goal of AI-driven efficiency is centered on digitizing human-made creations, such as accurately converting original artwork into 3D polygon models, reassuring fans that carelessly outputted AI data is not being included in released games.
Subsequently, Level-5 released a statement on September 17 specifically addressing the controversial elements in the artwork and PV for 'Yo-kai Watch 2: Haunted Domain'. The company clarified that unnatural power lines in the image were a designer's manual work simplified for the background based on real utility poles and wires, and that oddities with vending machines and rain gutters were human errors that occurred while simplifying details. Regarding a bicycle lacking brake levers, Level-5 explained that given its direction to enhance detail in the latest title, it should have been depicted this time.
Addressing observations that Keita's face differed between the PVs shown in the Nintendo Direct and Level-5's own showcase, the studio stated that this was not a change or modification caused by AI. According to Level-5, revisions were underway to improve the quality of the existing character model, but because the submission deadline for the Nintendo Direct video was earlier, those changes could not be included, whereas the revised model was used in its own showcase. Additionally, the company explicitly noted that no AI was used for 'Professor Layton and the New World of Steam'.
In its statement, Level-5 remarked, "By carelessly using AI in some videos, we created a situation where even human errors were misunderstood as AI usage. We deeply apologize for causing distrust due to our clumsy application of AI and human mistakes." The company added, "We will exercise greater care with public releases to prevent situations that cause misunderstandings," concluding, "Going forward, our company will limit AI use strictly to improving efficiency and avoid direct usage in videos presented to users."
