Civitai, a well-known platform for sharing AI models and a key player in AI-generated adult content, announced it will ban certain types of content following pressure from payment processors.
This includes content related to incest, self-harm, and fetishes involving bodily fluids like urine, vomit, menstruation, and diapers. Additionally, the site will restrict access to content that features real people and will make it harder to find such material.
In a recent post, Civitai explained the changes, stating that the scrutiny of AI content is increasing from various authorities and the public.
They acknowledged that platforms like theirs need to adhere to stricter standards regarding safety, legality, and accountability.
While Civitai has aimed to be a space for creativity and innovation, they recognize that some adjustments are necessary to continue operating and to protect their users.
Justin Maier, CEO of Civitai, shared that the original goal was to allow any legal and consensual content. However, he noted that the financial landscape for payment processing is more restrictive than what the law requires.
Several partners have made it clear that certain changes must be made to keep payment services active. To avoid disruptions that would negatively impact creators, Civitai decided to comply with these demands.
Although Civitai claims to be meeting higher standards, the content they are banning has often skirted the edges of what is acceptable on adult sites.
For example, diaper-related content is particularly sensitive because it can be misconstrued as involving minors. In 2023, Patreon banned numerous accounts producing “adult baby/diaper lover” content, which had been popular in the furry community.
Content depicting incest is generally avoided by most adult platforms due to its illegality in many areas, which is why many adult films use the common trope of step-sibling relationships instead.
Additionally, content involving weapons, self-harm, or certain bodily fluids is typically banned on adult sites. For instance, OnlyFans has similar prohibitions in place.
Civitai will give users a 30-day notice before enforcing the removal of content that violates these new rules. A quick search on Civitai reveals that it currently hosts numerous AI image generation models aimed at creating explicit content, including those related to firearms and bodily fluids.
While Civitai tries to downplay its role in adult content, the recent policy changes indicate that it faces the same moderation issues as other adult platforms, all of which are under pressure from payment processors.
This pressure has led to significant changes in content moderation practices across the industry, exemplified by P*rnhub’s decision to ban millions of unverified uploads in 2020.
Civitai’s moderation may be particularly challenging due to the nature of AI image generators, which can produce an endless variety of images based on user prompts.
This has led the platform to require users to include metadata for all adult content, ensuring that there is verifiable information for age verification and consent audits.
Maier explained that this metadata requirement is intended to align with federal regulations regarding record-keeping for adult content. Users can still keep their prompts hidden if they choose, but the generation details will be embedded for transparency.
Some users expressed mixed feelings about the new policies in Civitai’s Discord community. While some welcomed the reduction in certain types of content, others were concerned that this could lead to broader censorship of themes they enjoy.
Civitai is also taking steps to address models that recreate the likeness of real individuals. Content that includes the names of real celebrities will be hidden, and advertisements will not be shown alongside this material.
Users will also be prohibited from tipping for this type of content, which is a notable shift from previous practices that encouraged such creations.
This isn’t the first time Civitai has revised its content moderation policies. Following an investigation in 2023 that revealed users were generating potentially illegal images, the platform introduced significant measures to prevent abuse.
Other Stories You May Like