Custom Safeguards / Request to Disable Default Safeguards

Introduction:

Stability AI offers advanced AI image generation capabilities while maintaining adherence to legal and ethical standards. To ensure a safe and inclusive environment, we have implemented safeguards that prevent the generation of illegal or inappropriate content. In some cases, customers may wish to disable these safeguards if they have implemented their own content moderation practices. This article explains our custom safeguards evaluation process and requirements for eligible customers.

Custom Safeguards Evaluation Process:

To accommodate customers with specific content moderation needs, Stability AI offers the option to disable our default safeguards. However, we require customers to demonstrate that they can prevent the generation of violative content independently.

Eligibility Criteria:

Customers interested in disabling our safeguards must provide information about their existing content moderation practices and safeguards. This includes:

1. A detailed description of their content moderation practices, outlining how they prevent illegal or inappropriate content generation.
2. Evidence of effective implementation, such as case studies, demos, or certifications demonstrating their ability to maintain a safe and inclusive environment.
3. Signed agreements acknowledging responsibility for moderating content generated using Stability AI's services. (Provided by Stability AI upon eligibility approval)

Evaluation Process: 

Upon receiving the required information, our team evaluates the customer's content moderation practices and safeguards. If deemed satisfactory, we grant permission to disable our default safeguards. 

Conclusion: 

At Stability AI, we strive to provide flexible solutions while maintaining a commitment to safety and inclusivity. Our custom safeguards evaluation process ensures that customers who choose to disable our default safeguards can maintain a safe and ethical environment for their users. 

If you have specific content moderation needs, please contact us by submitting a Support Request to discuss eligibility requirements and evaluation procedures.