Introduction to Policies

In Clavata.ai, a Policy defines what is unacceptable on your platform. Policies guide both your platform’s users and moderation teams, empowering automated systems to enforce these rules, ensuring a safer and more structured environment.

This guide will walk you through some general processes to help in creating effective Policies that align with your platform’s specific use cases.

How to Approach Writing a Policy

Writing a Policy in Clavata.ai begins with understanding what types of content you want to filter out. Here’s how to get started:

1. Start by identifying the different types of user-generated content on your platform that need oversight. Examples include:
   • Usernames
   • Profile pictures
   • Private and public chat
   • Public forums
   • AI-generated inputs and outputs
2. Identify the contexts that content needs to be analyzed for.
   • A good starting point is reviewing your platform's terms of use, community guidelines, or similar user-facing guidelines. These documents provide the foundational framework for defining what content and behaviors are considered benign or harmful, and your Policies should mirror that.
3. Consider how to manage content that isn’t clearly benign or harmful. Defining how to handle ambiguous or contextual content is critical to developing well-rounded Policies. 
4. Your platform may allow for exceptions based on the context of the content. Specify when Policies should be overridden, such as when distinguishing between satire and harmful misinformation.
5. Consider how flagged content will be handled based on your Policy’s purpose. For example:
   • If your Policy will be determining what is publicly visible on your platform, you may opt for a higher-level approach, without needing to cover every possible scenario in detail.
   • If your Policy will be used for automating moderation actions against users, it’s crucial to be highly specific, particularly with gray-area content, to ensure precise and accurate handling.

Writing, testing, deploying, and monitoring your Policy is an iterative process. With each adjustment, your Policy becomes more refined, ensuring content safety evolves alongside your platform. Once your Policies are deployed, periodically review flagged content and update your Policies to address emerging issues.

Key Considerations When Creating Your Policy

When writing Policies for your platform, it's important to address the different types of content and contexts where violations may occur. Here are a few things to keep in mind while creating your Policies.

Content-Type

While Policies can handle both text and image content, different types or sources of content on your platform may require varying levels of oversight, which could benefit from distinct Policies tailored to those specific needs.

Examples:

• Usernames and Profile Pictures: This information is often visible platform-wide and can be displayed as suggested content, and can have a big impact on user experience. Consider a more fine-tuned approach to this kind of public-facing content.
• Comments and Posts: Text-based content should have flexible Rules based on context. For example, posts in public forums may require stricter analysis compared to direct messages in private conversations.

Areas of your Platform

It’s a good idea to start with a general Policy that applies to most of your content types, then refine it for more specific use cases. 

Example:

• You might start with a general Policy for all user content, then create a more refined version specifically for reviewing highly visible content like usernames or profile pictures.

Exceptions

Clavata.ai’s Policies allow for exceptions when necessary. When writing Policies, think about where exceptions should be applied. 

Example:

• If your platform is related to education, your Policy may flag sexual vulgarity, but you may want to have an exception for sexual terms used in a scientific or medical context. 

Need more guidance on writing Policies? Explore our additional resources or contact our support team at support@clavata.ai for more assistance.