AI is Ruining Game Sales, Numbers Show

The debate surrounding the implementation of artificial intelligence in the making of a game has moved from a distant tech-bro prophecy straight into the front lines of the video game industry.

Whether it is generating concept art, polishing voice lines, translating a few lines, or optimizing backend code, studios are increasingly experimenting with these tools to streamline production. As a result, seeing “AI-generated content disclosure” notices has quickly become a common sight for anyone browsing through digital store shelves to check out upcoming releases.

However, while the ethical conversation around human artistry has been buzzing for quite some time now, the narrative is surprisingly (or maybe not that much) shifting from a purely moral dilemma into a financial matter. That’s because a new study just revealed that relying on AI might be harming video games far beyond their creative essence.

Deciphering the Reality of Steam’s Numbers

According to a recent research performed by Ross Burton, PhD, Head of Product and Data at Game Oracle, the presence of an AI disclosure constantly acts as a red flag for most consumers. Burton discovered that approximately 21% of all games released on Steam in 2025 contained an official disclosure for some form of AI utilization, a number that has undoubtedly climbed since then.

Since Valve doesn’t publicly disclose exact copy sales, the industry relies on proxy methods to estimate commercial performance – more specifically, tracking the volume of user reviews a title receives post-launch. Game Oracle’s study of nearly 10,000 Steam releases between January and October 2025 painted a genuinely eye-opening picture: titles disclosing AI use averaged a measly four reviews in their first month compared to seven reviews for non-AI titles. Also, almost 20% of AI-assisted games received absolutely zero reviews, while their average review scores sat noticeably lower than their traditional counterparts.

To ensure outside variables weren't skewing the results, Burton constructed a rigorous causal statistical model, controlling for prior developer experience, publisher backing, genre, and release timing to compare identical scenarios. And the final calculation is quite staggering: games that disclosed the use of AI received about 53% fewer reviews than non-AI games. In a vacuum where two developers of equal talent and budget release a similar game, the traditional project would theoretically land 100 reviews while the AI-dependent version would fumble at just 47.

“They have talent, budget, and know-how,” Ross Burton explained regarding competent teams trying to optimize their workflows. “They decide to experiment with AI to optimize their workflow. If this narrative is true (if 'good' studios are using AI), then AI use is catastrophic, resulting in a 40% to 60% drop in sales.”

Interestingly, the study also noted that inexperienced indie creators with no marketing budget saw minimal negative impact from using AI, mostly because those projects were already destined to struggle to find an audience. Instead, the stigma strictly punishes established, competent developers who actually have a built-in fanbase and a legacy to lose. To sum it up, while smaller titles managed to find success despite utilizing the technology, the broader market shows prominent intellectual properties taking some reputational hits.

In the end, Burton notes that it might not just be direct consumer backlash driving these negative numbers, but rather that heavy AI reliance often correlates with poor design choices and lazy craftsmanship in people’s perspective.