Legal
Acceptable Use Policy
1. Scope
This policy applies to all xMAD.ai models, weights, hosted agents and evaluation harnesses, and to any derivative you create from them.
2. Prohibited uses
- Creating or distributing malware, or gaining unauthorised access to any system.
- Deploying an agent to take irreversible actions on production systems without human review.
- Generating content that sexualises minors, or that incites violence against a person or group.
- Producing disinformation intended to influence an election or a public health decision.
- Surveillance, profiling or scoring of individuals without their informed consent.
- Circumventing our safety evaluations, sandbox, or rate limits.
- Any use that violates applicable law, including export control and sanctions law.
3. High-risk domains
Use in medicine, law, employment decisions or critical infrastructure requires human review of every output and a documented escalation path. Tell us before you deploy in these areas.
4. Reporting
Report a suspected breach of this policy through the contact page. We may suspend access while we investigate.
5. Enforcement
Breaches may result in suspension of hosted access and termination of the relevant licence. Serious cases are referred to the competent authorities.