Capcom will not add generative AI to Resident Evil Requiem and new games

Capcom has reassured fans and some of its employees: the publisher will not implement AI-generated materials in Resident Evil Requiem, Pragmata, and other upcoming games. However, the company is not completely abandoning generative AI — it is considering it as an internal tool for accelerating development.

Where the Dispute Started

The recent demonstration of DLSS 5 sparked concerns, with Resident Evil Requiem appearing among Nvidia's partners. After that, according to industry sources, conversations began within Capcom about whether the company was changing course and planning to more actively integrate generative AI into final game content. For a studio with a rather strict stance on such technologies, this looked like a notable signal, and the reaction from employees and players was nervous.

This was especially the case given that Capcom was not previously associated with attempts to make neural networks part of the creative result. The situation was amplified by the overall climate around DLSS 5. The technology was already criticized for potentially changing the visual style of games and making the picture too "plastic" and generic.

Against this backdrop, many perceived Capcom's participation in the presentation not as ordinary technical partnership, but as a possible hint at broader AI integration into development. Therefore, the company's statement to investors essentially became an answer not only to the market, but also to its own team, which wanted to understand where the red line is drawn.

What Capcom Said

On March 23, in a published Q&A session for investors, Capcom formulated its position quite directly: materials created by generative AI will not be implemented in game content. This is the key part of the entire story, because it's precisely what separates the company's current approach from the scenario that critics feared — when neural network assets start directly entering game releases and affecting their artistic appearance.

"We will not implement materials created by generative AI in game content."

At the same time, Capcom is not opposed to the technology as such. The company specifically emphasized that generative AI can be useful as a supportive tool for improving efficiency and productivity within the production process. In other words, this is not about replacing artists, composers, or programmers with neural networks, but about trying to accelerate routine stages, reduce time spent on some internal tasks, and free up teams where it does not directly affect the final authorial result.

Where AI Is Being Used

Currently, Capcom is examining exactly which zones generative AI actually provides speed advantages and does not compromise quality. The company directly lists several areas where the technology is considered not as a source of finished content for the player, but as part of an internal pipeline. That is, the focus has shifted from public results to behind-the-scenes processes, which usually remain invisible to audiences but significantly impact production timelines and costs.

• preliminary graphic materials and service documentation
• assistance with sound work and individual production tasks
• support for programmers in technical and routine stages
• idea generation at early stages of world design and mechanics

Capcom has already tested this approach before. According to the company, it built a prototype AI system based on Google Cloud that helps quickly develop ideas for game world design and basic mechanics. This does not mean automatic game assembly from a text prompt, but shows where the industry is primarily moving generative AI: not into final assets, but into preproduction, rough preparation, and internal organization of development. Essentially, the publisher is trying to take from the model the role of accelerator, but not the role of co-author.

What It Means

Capcom has chosen a cautious middle ground: no neural network materials in the game itself, but active experimentation with AI within the studio. For the market, this is an important signal: even major publishers who are not ready to hand over the creative part to generative models still see in them a way to reduce costs and speed up individual production stages. It seems that precisely such a compromise will become the most realistic scenario for the gaming industry in the coming years.