ArtStation Adds AI Declarations and Community Reporting Filters

ArtStation now requires every upload to carry a generative-AI declaration, and by default the platform hides any project declared as AI-generated. The update, announced following KitBash's acquisition of the platform, extends a policy that began in September, when ArtStation enabled the 'NoAI' tag on uploads by default and marked them as unavailable for AI training.

The new system rests on two mechanisms: creator self-declaration and community reporting. Publishers must classify each piece as primarily generative AI, a mix of generative and non-generative methods, or 100% non-generative.

ArtStation says most users declare their work honestly. The platform also acknowledges a persistent problem: creators uploading AI-generated content without flagging it.

What does the update change for artists?

To close that gap, ArtStation has opened a dedicated reporting channel. Users who encounter unlabelled AI work can report it, and the platform says every report feeds its AI filtering for the whole community.

Filtering itself now runs on three user-selectable levels:

• Strict — hides artwork when only a few reports suggest it may be AI-generated
• Balanced — filters content at a moderate level of reports
• Loose — removes work only when user signals are strongest

Users can also choose whether to hide all declared generative AI content or allow mixed generative and non-generative workflows. Out of the box, the default configuration hides any project declared as AI-generated, tagged 'CreatedwithAI', tagged with generative AI tools, or suspected as AI — with the suspicion filter set to Strict.

The stakes for the platform are concrete. ArtStation risks becoming unusable if low-quality AI-generated content buries work that users actually want to find, and the volume problem is the reason KitBash pledged to address user concerns when it bought the site.

What are the risks of user-led moderation?

The reception among artists has been mixed. Many users welcomed the update as a step in the right direction; others doubt its effectiveness.

The system depends heavily on users accurately identifying AI art. Some unlabelled AI work will likely go undetected, while the Strict setting could remove work falsely identified as AI. ArtStation has not stated whether artists will be notified when their work is flagged, or whether they can contest the decision.

That gap matters. Established digital artists have already been falsely accused of using AI elsewhere — the illustrator Loish has faced such accusations on TikTok — so false positives on ArtStation could penalise human-made work.

Why a full ban appears off the table

The policy also reflects commercial reality. ArtStation has a strong focus on concept art, and parts of the industry now treat generative AI as an inevitable component of the concept pipeline. The recent Disney deal with Adobe targets custom AI models developed specifically for concept work.

In that context, KitBash likely views a blanket ban on AI art as counterproductive, if not impossible to enforce. The declaration-plus-filtering model is the compromise: it leaves AI-assisted work on the platform while giving users granular control over how much of it they see.

What happens next depends on two unanswered questions — whether reporting accuracy holds at scale, and whether ArtStation adds an appeals path for artists caught in false positives.