itch.io AI Game Policies in 2026: A Practical Guide to Disclosure, Assets, and Store Pages

Context: Why itch.io's AI Rules Matter More Than Ever

itch.io has become the default home for experimental, small-team, and solo-developed games. That makes its AI policies unusually consequential. When a platform hosts hundreds of thousands of projects — from game jams to commercial releases — its rules on AI-generated content shape what independent developers can ship and how they describe it.

The pressure has been building since 2023, when AI art and writing tools moved from novelty to default workflow for many small studios. itch.io's response has been iterative rather than dramatic: the platform has leaned on creator disclosure and community moderation rather than blanket bans. That approach contrasts with Steam, which since 2024 has required developers to disclose AI-generated content during the store-page setup process and separates "pre-generated" from "live-generated" content. It also contrasts with some asset marketplaces that banned AI assets outright before reversing or refining those bans.

For developers, the practical question is no longer "is AI allowed on itch.io?" It mostly is, with conditions. The real question is: what do you need to disclose, where, and how do you do it without scaring off players who are wary of AI slop? This guide treats the policy as a product decision, not just a compliance checkbox.

Core Analysis: What itch.io Actually Expects

Disclosure is the load-bearing rule

itch.io's model is transparency-first. Creators are expected to tell players when a game contains AI-generated content, and to do so in a way that is visible on the project page. The platform has not published a single rigid taxonomy the way some storefronts have, but the operational expectation is consistent: if a meaningful portion of your art, audio, code, or writing was produced by a generative model, say so.

The ambiguity is deliberate. itch.io is a hosting platform with a strong anti-censorship ethos, and its moderation tends to react to complaints and patterns rather than pre-approving every upload. That means enforcement is uneven — but the reputational risk is not. A player who discovers undisclosed AI art in a paid game is far more likely to leave a public complaint than a private refund request.

The asset question is where most developers trip

Most disputes on itch.io are not about whether a game used AI to brainstorm names. They are about assets: character portraits, backgrounds, music, voice lines, and 3D models. The platform's expectation is that AI-generated assets are disclosed in the project description or a dedicated credits section. If you used a tool like Midjourney for concept art that was later painted over, the honest answer is usually "AI-assisted" rather than "AI-generated" — but you should still mention it.

The same logic applies to audio. If you generated a soundtrack with Suno or Udio, disclose it. If you used Play.ht for placeholder voice lines that were later replaced by human actors, that is worth a line in your devlog but not necessarily a store-page warning. The threshold is whether the AI output survives into the shipped build.

Store-page labeling is a design problem, not just a legal one

A disclosure label can be a conversion killer if it reads like an apology. The better pattern is to make it specific and confident: "Background art generated with Stable Diffusion and refined in Photoshop; all character art hand-painted." Specificity signals craft. Vague labels like "some AI used" read as evasive and invite suspicion.

This is where itch.io's format helps. Project pages support rich text, devlogs, and custom sections. You can put a short "How this was made" block near the top of the description, then link to a longer devlog for readers who want detail. That structure satisfies disclosure while turning it into a trust signal.

Real Cases: What the Data and Precedents Show

Steam's AI disclosure requirement, introduced in January 2024, is the clearest precedent for how storefronts are handling this. Valve's approach asks developers to categorize AI use into pre-generated content (created during development) and live-generated content (created at runtime), and to describe the guardrails for the latter. itch.io has not adopted an identical form, but the categories are a useful mental model for itch.io creators writing their own disclosures.

On the asset side, the history of AI policy on marketplaces has been volatile. In 2023, some stock and asset platforms moved to ban AI-generated uploads, then walked those bans back within months as the technology became embedded in professional workflows. The lesson for itch.io developers is that platform rules are still settling, and a disclosure you write today may need revision in six months. Keep a plain-text record of which tools you used, at which version, for which assets.

Within the itch.io ecosystem, the most visible friction has come from game jams and bundles. Several high-profile jams have added their own AI rules on top of the platform's — some banning AI art entirely, others requiring disclosure in the submission form. If you are entering a jam, read the jam rules first; they are often stricter than itch.io's baseline.

Practical Advice: A Compliance Workflow That Doesn't Hurt Your Page

• Inventory before you upload. List every asset category — art, audio, code, writing, 3D — and mark each as human-made, AI-assisted, or AI-generated. This takes twenty minutes and prevents most disputes.
• Write a specific disclosure block. Name the tools and the scope. "Music generated with Soundraw, mixed in Reaper" is better than "AI music used." Place it in the project description, not buried in a devlog.
• Separate pre-generated from runtime AI. If your game calls an LLM at runtime — for dialogue, for example — say so explicitly and describe what the player will see. This is the category most likely to draw scrutiny.
• Keep receipts. Save tool names, versions, and dates for each asset. If a policy changes or a complaint arrives, you can respond with facts rather than memory.
• Use devlogs as a trust channel. A short post explaining your AI workflow — what you kept, what you rejected — turns a compliance requirement into content that players actually read.
• Check jam and bundle rules separately. Platform policy is the floor, not the ceiling. Community events often set stricter terms.
• Audit before major updates. If you swap in new AI-generated assets after launch, update your disclosure. Stale labels are a common source of complaints.

Outlook: Rules Will Tighten, Disclosure Will Normalize

The direction of travel is clear. Storefronts are moving from binary bans toward structured disclosure, and itch.io is likely to refine its guidance as AI-generated games become a larger share of uploads. The platform's community-driven moderation means the practical enforcement will continue to come from players and jam organizers as much as from staff.

For developers, the strategic move is to treat disclosure as part of the craft narrative rather than a legal formality. The studios that will fare best are the ones that can say, precisely and without defensiveness, what the machine did and what the human did. That is a harder sentence to write than "AI-free," but it is the one that will hold up as the rules keep moving.