Did LEVEL-5 use AI in its games? Hino’s response explained

LEVEL-5 president and CEO Akihiro Hino has confirmed that AI was used in footage shown during LEVEL5 VISION 2026 II Dream. He apologised to people made uncomfortable by that presentation choice, but he did not promise that AI is absent from LEVEL-5’s game-development pipeline.

The narrower answer is that Hino defended continued AI-assisted efficiency work while drawing a boundary around what he described as the human-made creative core of LEVEL-5’s released works. He also said released games would not contain “simplistic AI-generated output data.” That is a company assurance with an undefined threshold, not independent verification of every asset or production process.

Hino connected the company’s AI work to an effort to shorten development of major titles from five years to two. Nintendo Life’s contemporaneous English rendering of his statement reports both that efficiency target and the limits he says LEVEL-5 places on AI-assisted work.

What LEVEL-5 confirmed about the presentation

LEVEL-5 broadcast LEVEL5 VISION 2026 II Dream on 10 September 2026. Two days later, Hino acknowledged that AI had been used in footage shown during the event. According to Nintendo Life’s reproduction of his statement, he said the intention was to make the presentation more spectacular and apologised to those who found the use uncomfortable.

This confirms AI use in the presentation footage. It does not, by itself, establish that a particular trailer, character, game asset or upcoming title contains AI-generated production material. Suspected visual artefacts identified in online discussion are not a substitute for technical documentation or a title-specific company disclosure.

Hino also discussed LEVEL-5’s broader production plans. He said the company is seeking AI-assisted efficiency mainly when digitising creative work made by people, giving the conversion of original artwork into polygons as an example. He said scenarios, character designs and basic settings—the elements he associated with a work’s charm and individuality—are created by people. Those claims are set out in Nintendo Life’s translated account.

What Hino said about released games

The most important assurance in the statement is also carefully limited. In Nintendo Life’s English rendering, Hino said:

“Please rest assured that no simplistic AI-generated output data is included in the games that are released.”

Nintendo Life

Game*Spark’s report on Hino’s response gives the key Japanese phrase as 安易なAI出力データ. The English wording above should therefore be read as an attributed translation rather than a technical definition supplied by LEVEL-5.

“No simplistic AI-generated output data” is not equivalent to “no AI is used anywhere in making LEVEL-5 games.” The statement does not define what makes output “simplistic,” provide an exhaustive list of covered workflows or explain whether generated material might be altered, rebuilt or discarded before release.

It also does not specify how the assurance applies across visual art, 3D models, animation, audio, text, code, localisation and other production materials. Readers can treat it as Hino’s stated boundary for released games, but not as evidence that every asset has undergone an independent audit.

Why LEVEL-5 is continuing its AI-efficiency work

Hino’s apology did not amount to a rejection of AI-assisted production. He tied the work to an ambitious development-speed target and described the company as still experimenting:

“Ultimately, under the plan to shorten the current 5-year development period for major titles to 2 years, we are in the stage of trial and error, and struggling forward.”

Nintendo Life

This is a target, not a claim that LEVEL-5 has already reduced major-game development to two years or a guarantee that it will achieve that schedule. It explains why Hino apologised for the presentation while continuing to defend AI as a potential production-efficiency tool.

Other publishers have drawn comparable distinctions between efficiency tools and human control over core creative decisions. Capcom’s generative-AI policy offers relevant industry context, although it does not provide evidence about LEVEL-5’s practices.

What LEVEL-5 still has not disclosed

Hino’s statement provides principles and an example rather than a complete account of LEVEL-5’s production pipeline. The available text does not identify the AI systems, model providers or vendors involved. It does not explain the provenance of training data, whether particular tools rely exclusively on owned or licensed materials, or what contractual safeguards apply.

The statement also does not provide a title-by-title inventory of AI use, identify every LEVEL5 VISION 2026 II segment that used AI or define the scope of “mainly” when describing the digitisation of human-created work. It therefore leaves open which other workflows, if any, are covered by the company’s efficiency programme.

These omissions do not prove misconduct or establish that AI-generated output appears in a released LEVEL-5 game. They mean only that the public cannot determine the full scope of the company’s practices from this statement. It is the same methodological distinction between a company assurance and verifiable implementation explored in our guide to what an AI disclosure does—and does not establish.

The clearest conclusion is that Hino confirmed AI use in LEVEL-5’s September presentation footage and described an ongoing effort to use AI-assisted tools mainly to digitise human-made creative work. He also said released games would not contain what the translated statement calls “simplistic AI-generated output data.” LEVEL-5 has not supplied enough technical, contractual or title-specific information to turn that assurance into a complete account of AI use across its games.