CivitAI Cracks Down on Deepfakes: Visa, Mastercard Force Policy Changes

CivitAI, a leading AI model repository, has significantly revised its policies on NSFW content, particularly regarding celebrity LoRAs, under pressure from payment processors Mastercard and Visa. These changes, described by CivitAI's Community Engagement Manager as unwanted but necessary to avoid legal issues, include hiding content tagged with real people's names from public feeds and filtering out content with child/minor themes. The platform now partners with Clavata AI for enhanced moderation, aiming to prevent the re-upload of removed images depicting real individuals, even if unseen before. This action comes after several high-profile individuals, including actors and actresses, requested the removal of their likenesses. The changes also involve mandatory metadata for all images and videos to prove synthetic origin, a minimum 50% noise alteration for BYOI images to avoid subtle deepfakes, and the disabling of tipping and Early Access for real-person models. While CivitAI attempted to mitigate the impact, they were left with limited options. The changes have raised concerns among users, prompting initiatives to preserve removed LoRAs. The pressure from payment processors highlights the growing legal concerns surrounding AI-generated NSFW content and deepfakes, pushing the platform toward stricter moderation and potentially driving similar changes across other AI platforms. The necessity of these measures underscores the evolving legal landscape and the challenges faced by platforms hosting user-generated AI content. The long-term impact on the platform's user base and its ability to retain its popularity in the face of these restrictions remains to be seen.

The growing concerns over ai automation deepfakes have prompted major payment processors to pressure platforms into implementing stricter content moderation policies.

The payment processors' pressure on CivitAI reflects broader concerns about chatgpt automation deepfakes and their potential misuse across various platforms.

(Source: https://www.unite.ai/civitai-tightens-deepfake-rules-under-pressure-from-mastercard-and-visa/)