Right after IGF Judges Announced They Will Not Vote for Any Games Using Generative AI, *Big Walk* Admits It Used AI Tools in Development

Judges for IGF (Independent Games Festival) first released a statement: the number of entries developed with generative AI this year has increased noticeably, and they will not vote for any game that uses generative AI "to any extent". The statement acknowledged that entries for this year's IGF are all excellent, but discussions around AI among judges were extremely heated and no universal consensus was reached.

Shortly after, the development team of indie game Big Walk also released a statement: they used AI "sparingly" during the final stretch of development to help handle tricky technical issues and bugs. The original text uses the exact term "sparingly", and specifies the usage took place in the "final stretch" of production. Putting the two announcements together, the specific dispute becomes clear: judges want to define whether a game used AI at all, while developers are focusing the conversation on how AI was used and for what scenarios.

This is also the most unavoidable sticking point now that generative AI has entered game development workflows. When coding, using an AI model to look up an error, explain a log entry, or find clues for a bug — does that count as AI use "to any extent"? If it counts, how will the jury verify such use? If it does not count, who gets to draw the clear boundary of allowed use? The IGF judges' statement made their stance clear, but did not provide actionable judgment criteria. Big Walk's response also did not elaborate on details, only confirming the usage scenarios and development phase.

The Verge's report laid out the clear timeline: the judges' statement came first, followed by the development team's response. Neither side denies that AI has already been integrated into independent game development processes. The current conflict is that judging rules are still stuck at the level of public stance-taking, while actual development practice has already incorporated AI into granular tasks like a single debug session or a specific technical challenge. Whether the award will exclude entries for such usage will become a far more practical issue to address going forward.

Original report: The Verge