Usage Policy

We build AI to benefit humanity. This policy defines the boundary between safe discovery and dangerous misuse.

1. Universal Prohibitions

You may not use Metanthropic models for:

• CSAM: Generation or promotion of Child Sexual Abuse Material.
• Violence & Hate: Content that incites violence or promotes hate speech against protected groups.
• Malware Creation: Generating code specifically designed to disrupt, damage, or gain unauthorized access to computer systems.
• Political Campaigning: Generating high-volume propaganda, targeted campaign materials, or lobbying content.

2. High-Risk Use Cases

Certain applications require higher standards of verification ("Human-in-the-Loop"). You must not deploy our models in these domains without our express approval:

3. The "Glass Box" Requirement

If you are using our models for scientific research or public dissemination, you must disclose that the content is AI-generated. Furthermore, you are encouraged to publish the model's "Thought Trace" alongside the output to allow for public verification of the result.

4. Enforcement

Our safety systems monitor input/output patterns. Violation of this policy will result in immediate API key revocation. We cooperate with law enforcement in cases of illegal activity.