The developers believe the accusations regarding their use of artificial intelligence are unfair.

The widespread presence of generative artificial intelligence is difficult to digest, and it is even harder to determine what truly qualifies as AI-generated content and what does not. Today, anything that merely looks like it could have been created by AI can easily become a target of criticism, especially as some games openly claim to be fully AI-generated.

Positive Concept Games, the developer behind the SNES-style RPG Shrine’s Legacy, shared a Steam review on Twitter that labeled the game as AI trash, described its story as “dog sh*t mixed with cat sh*t”, and repeatedly claimed it was made using ChatGPT. The developer urged people not to do this, explaining that the team spent years working on the game and collaborated exclusively with real human artists across every aspect of development. They do not support generative AI and state that they will never use it.

Please don’t do this.
We poured years of our lives into this game and only worked with real human artists on everything: From the writing to the coding, all work was done by human hands. We do not endorse generative AI and will never use it. pic.twitter.com/3L7NKVX1L8

— Shrine’s Legacy – Out Now! (@ShrinesLegacy) December 10, 2025

Other Steam reviews examined the accusations more closely. One review shared an old tweet from the developer showing the use of an AI tool as part of an April Fools’ joke, although the related artwork does not appear in the game and the joke itself was centered on AI usage. Another review stated that the game was purchased not to play it, but to understand the controversy and look for perceived inconsistencies, such as minor anomalies in character art or a feeling that the story was AI-generated. The developer responded by claiming this user was almost certainly an alternate account of the original accuser.

Even major studios such as Ubisoft have adopted an apologetic rather than permissive stance when disclosing AI-related information. As AI becomes increasingly common and harder to identify, accusations like these can gain significant traction even without evidence. According to a Microsoft study, people correctly identify AI-generated content only 62% of the time, which, while more accurate than AI-generated search results, still adds an unsettling ambiguity to the debate.

Source: PCGamer, Steam, Steam, Ars Technica