Are AI-Generated Video Games on the Way?

AI-generated video games are becoming an increasingly discussed topic in the gaming industry as new technology continues to emerge.

Microsoft’s latest innovation, Muse, is a tool that claims to streamline the design process using AI-generated gameplay videos.

While Muse introduces a novel way for developers to test ideas efficiently, it has also sparked debate about its true capabilities.

Some experts believe it could be the first step toward true AI-generated video games, while others argue it offers little practical value.

As the gaming industry rapidly evolves, Muse’s potential impact remains a topic of considerable interest.

What is Muse and How Does it Work?

In early 2025, Microsoft introduced Muse, described as the world’s first World and Human Action Model (WHAM).

Developed in collaboration with Ninja Theory, Muse has been trained on tens of thousands of hours of gameplay data from Bleeding Edge.

By learning from this extensive data set, Muse can generate realistic-looking gameplay clips that designers can manipulate using prompts.

Microsoft claims this will help developers test ideas more efficiently without committing resources to full-scale implementation in traditional game engines.

The tool is designed to help developers visualise ideas without investing significant time in building them in a gameplay engine.

If a designer wants to explore how a power-up might affect gameplay, for instance, Muse can generate a mock video showing its potential impact.

Julian Togelius, associate professor in computer science and engineering at the New York University Tandon School of Engineering, said:

“Game engines are complicated, messy things and it takes a lot of time to simulate things – they’re not built for that.”

“[Working with] a simulation of the game can be much easier and faster.

“The opportunities opened up by this kind of study are pretty big, but the limitations are also real.”

Limitations & Concerns

Despite its innovative features, Muse cannot create entirely new games or generate playable simulations.

Instead, it produces visual mock-ups based on the data it has been trained on.

As Togelius explained: “This particular model is trained on 500,000 game sessions, so likely around 100,000 hours of gameplay. But it only works because you have so much data.

“If you move far beyond what’s been recorded, simulations generally stop behaving well.”

Muse’s reliance on extensive gameplay data makes it more suitable for live-service games such as Bleeding Edge.

For smaller or single-player games, the effort required to train an AI model like Muse may be excessive and impractical.

Ken Noland, a veteran game designer and founder of the AI-focused co-development company AI Guys, expressed doubts about Muse’s practical value.

He said: “It’s an amazing technical hurdle that they’ve jumped, but it kind of feels like they’re going through their Zoom moment: a product coming into a market that doesn’t really have a purpose.

“The technology is cool, and don’t get me wrong, video generation is not an easy thing to do… I just don’t see its target audience.

“Game developers won’t be able to use it for rapid production because it doesn’t actually, aside from visualising a particular thing, address any underlying game development issues.”

The Unclear Future of AI-Generated Games

Adding to the confusion surrounding Muse’s capabilities, Microsoft Gaming CEO Phil Spencer claimed the tool could aid in preserving classic games.

He implied that Muse’s AI models could “learn” old titles and emulate them on modern hardware.

Microsoft CEO Satya Nadella added to the speculation by suggesting that Muse was the first step toward building a “catalogue” of AI-generated games.

However, there is currently no clear evidence that Muse can achieve this.

Togelius said:

“I will choose to graciously interpret what Satya said as visions of what could be done in the future.”

“It’s entirely possible that we will get to some version of that, but it’s not around the corner. What Microsoft has done in this paper is a foundation stone.”

While Muse has sparked interest, it does not yet create fully functional, AI-generated video games.

Its primary role remains focused on generating visual concepts and assisting designers during the ideation phase.

Other AI Innovations in Gaming

Muse is not the first AI-driven innovation in gaming.

In 2024, Google launched GameNGen, a playable version of Doom that functioned without a traditional game engine.

While initially successful, Google’s model struggled with consistency, generating inaccurate game elements as play sessions continued.

More recently, Google released Genie 2, which claims to generate “playable worlds”.

Although promising, Genie 2 remains in the early stages and has yet to demonstrate the practical reliability developers require.

AI-generated video games are still some distance from becoming a mainstream reality.

Microsoft’s Muse introduces a novel way for developers to test ideas and visualise gameplay changes quickly.

However, Muse’s practical use in full-scale game creation remains uncertain.

As the technology matures, developers may find new ways to blend AI capabilities with traditional design methods.

For now, Muse stands as a significant technical achievement but far from the revolution some have predicted.