Nexus Mods makes AI disclosure mandatory for new uploads and rules out an AI ban

No AI model is involved. The change is a platform governance policy.

• Uploads: authors must declare AI use. AI-assisted mods need a public explanation of how and where AI was used (The Verge).
• Mod pages: AI tags sit right after adult-content tags (Neowin, via search summary).
• Users: controls to hide AI content on game pages and the homepage, with search filtering "next up" (The Verge).
• Feed: caps on how many mods from one author appear at once in "Recently Added," and tighter limits on uploads from new accounts.

• Taxonomy:
• AI-Generated Content: AI-made code, UI, voices, dialogue, translations, music or in-game assets.
• AI Assisted: human-led work with limited AI help, such as texture or audio upscaling, or concept art that is not in the final mod.
• AI Media: AI-made thumbnails, videos or descriptions only.

Source: Shacknews.

• Provenance detail: for AI-assisted mods, Inven reports that authors must name the tools and models used and say where their outputs appear in the final mod (Inven Global).
• Verification: moderators review manually and users are asked to report non-compliant mods (VGC). No automated detection or accuracy figures have been published.

• A ban would drive AI use underground. This is Nexus's own rationale, not tested evidence (The Verge).
• No training on mods, and no selling or licensing them for training. This is a company pledge with no independent audit. Inven also reports a planned permission setting that would require explicit creator consent before mods are used for AI training (Inven Global). The Verge describes it more vaguely, as "clearer control" over reuse. The two accounts differ (The Verge).
• Bulk uploads are increasing. Folmann says the problem is "less of an issue related to AI" and comes from faster mod production. Nexus has published no upload figures.

• In 2023 Nexus removed AI content only after complaints. Tim Friedlander, president of the National Association of Voice Actors (NAVA), sent the site a list of suspect AI mods, and some were taken down (Kotaku).
• The July 2026 tags were the first step from removal-on-complaint toward required labeling (Shacknews).
• The Verge places the move alongside Valve's response to a "significant increase" in Steam game submissions, which points to a wider problem for content platforms (The Verge).

• The community is split. Some users want AI banned outright and are tired of AI thumbnails. Others say AI coding tools have made modding more accessible and feel singled out (VGC). PC Gamer's headline says Nexus is still "playing softball with the tide of slop" (PC Gamer).
• Enforcement depends on honesty. Disclosure is self-reported and violations are caught through manual reports, so undisclosed AI use may go undetected.
• Training controls are inconsistent across reports. Outlets describe the training-consent controls differently (see above).

• UGC platforms: label, filter and keep a public record of how AI was used, rather than banning it. That is likely to become a reference model for user-generated content sites.
• Tool vendors: creators may have to name specific models in mod descriptions. Tools that log what they generate and where it ends up will make compliance easier.
• Mod authors: check existing uploads before January 2, 2027 to avoid warnings, restrictions or removal.
• Discovery: the controls to hide AI content create a separate audience segment. Mislabeled mods risk moderation, and correctly labeled ones may lose visibility among users who filter AI out.
• Data teams: treat modding sites as places where the license terms for AI training are being tightened.