Big Walk's 10 Lines of AI Code Spark IGF Jury Reckoning
By @sharedot · · 6 pages
- Culture
- Games
- AI
- Indie Games
House House disclosed that an LLM helped fix a Unity bug in Big Walk, and 19 IGF jurors now refuse to vote for any game touching generative AI.
Nineteen jurors draw a hard line
The fight over AI in indie game development broke into public view this week.
According to Ars Technica, 19 members of the 50-plus-member Independent Games Festival Awards jury released a signed statement 'declining to vote for any game that uses generative AI in any capacity,' citing a 'significant increase in the number of submitted games developed with generative AI tools.'
The group named no specific games, but pointed to AI's ecological, economic, psychological, and labor harms, arguing the technology's adoption is still 'not a foregone conclusion in our industry.'
The jurors also urged the IGF as a whole to adopt a strong anti-AI policy that disallows games containing any AI-generated work in future years.
Ten lines of LLM-assisted code
By Tuesday morning, one critical darling sat at the center of the debate.
House House, the studio behind the co-operative exploration game Big Walk, told Aftermath that 'LLMs were used sparingly' in the final stretch of development to work through difficult technical challenges and bugs, estimating that well under 0.1% of the game's code touched this usage, with generative AI used in no other context.
After a Unity engineer concluded the fix 'wasn't possible,' the studio posed the question to an LLM chatbot, which suggested a fix the developers modified extensively into roughly 10 lines of code inside URP.
Why the apology stunned fans
What makes the moment surprising is that House House disclosed the use at all — and apologized for it.
The studio said it made no disclosure on Steam because Valve does not request disclosure of code or chatbot usage, but posted the statement publicly now because, per Aftermath's reporting, 'it has become a matter of public interest.'
House House stressed the disclosure 'is not an endorsement of, or argument for, the use of AI tools, however minimal,' warned that voluntary disclosures might contribute to the technology's normalization, and called its lack of an internal AI usage policy 'a failure' it will correct going forward.
Ars Technica notes jurors including Austin Walker and Damon Reece had described the game as 'remarkable' and deeply special, yet said even minimal AI use overrode that respect for the work itself.
Zero tolerance or proportionality?
The community has split sharply over what happens next. Ars Technica reports Big Walk's Steam page has been hit with over 400 negative reviews since the disclosure, though anti-review-bomb tools are limiting damage to its overall 'Very Positive' reception.
Critics of the absolutist stance are speaking out: Ars Technica quotes judge Liz Ryerson arguing she didn't find the use disqualifying, saying it bothers her that 'someone gets punished a lot more for transparency about a smaller level of fuckup,' and Steam curator Dominic Tarason calling Big Walk 'very mildly compromised art' whose failing is 'that's lame, don't do that again.'
Developer Brendan Keogh told Ars Technica that extreme zero tolerance is 'ultimately counterproductive' and encourages hiding AI use, while narrative designer Hannah Nicklin defended zero tolerance as 'a viable political position.'
Precedent exists: Ars Technica notes the Indie Game Awards rescinded honors for Clair Obscur: Expedition 33 last year over AI-generated textures.