When KitBash bought ArtStation, it pledged the address users concerns about the flood of AI content on the digital art platform. It took the first step back in September, when it turned on 'NoAI' for uploads by default, marking them as unavailable for AI training.
Now a new update introduces an attempt to police generative AI content along with more complete AI declarations and more control over filtering. The aim is to make it easier to users to control what they see, but it's got a mixed response from artists so far.
We're introducing the next phase of generative AI content controls. Creators will have clearer AI declaration requirements when publishing, and AI filtering is now on by default, with options that give you finer control. Learn what's changing → https://t.co/cuHmcoAuhw https://t.co/64PevA0XjmNovember 9, 2026
When publishing content on ArtStation, users will now be asked to declare whether generative AI was used in the creation of the piece. Creators can identify their work as primarily generative AI, mixed generative AI and non-generative methods or 100% non-generative.
This relies on users being upfront, and ArtStation says that most are. However, it's recognised that there is a problem with people uploading AI-generated content without flagging it as such.
That leads to the second aspect of the latest update, which is a dedicated channel for reporting AI content. If users come across artwork they think is AI-generated but which isn't labelled, they can report it. ArtStation says that every report will help improve its AI filtering for the whole community.
Users also get more control over what they're shown: they can choose whether to hide declared generative AI content, or to allow mixed generative and non-generative workflows, and they can adjust how aggressively ArtStation filters content based on user reports of suspected generative AI.
The 'Strict' setting hides artwork when just a few reports suggest it may be AI-generated. 'Balanced' filters it out when there’s a moderate level of reports and 'Loose' removes it only when user signals are strongest. By default, the filter will be set to not show any project declared as AI-generated, tagged with 'CreatedwithAI' or tagged with generative AI tools, and suspected as AI will be set to strict.
The platform risks becoming useless if it's impossible to find interesting work because of the sheer volume of low-quality AI-generated content. Many users have welcomed the latest update as a move in the right direction, while others have raised doubts about how effective the measures it will be.
There's a strong reliance on user moderation, which will depend on users accurately identifying AI art. Some may still go undetected completely, while the 'Strict' filtering could remove work falsely identified as AI.
It's not yet clear if artists will be notified if their work is flagged or whether they will have a way to contest that decision. That could be a concern when even long established digital artists like Loish have been falsely accused of using AI over on platforms like TikTok.
The update highlights the difficult balancing act of policing AI content. ArtStation has a big focus on concept art, and some in the industry now see generative AI as an inevitable part of the pipeline in this field. The new Disney deal with Adobe will focus on developing custom AI models specifically for concept work. In that context, KitBash probably sees banning AI art from the platform as counterproductive if not impossible.
