EN
Open developer docs

ACCEPTABLE USE

Acceptable Use Policy

This policy applies to all use of Octoryn Router. It supplements the Terms and customer agreement. Customers are responsible for their users, applications, prompts, outputs, tool execution and downstream actions.

1. Lawful and authorised use

You must use the service only for lawful, authorised purposes within the approved workspace and service scope.

  • Do not violate privacy, intellectual property, consumer protection, anti-discrimination or other applicable rights.
  • Do not submit data or content you lack authority to process.
  • Do not misrepresent model output as verified human or professional advice.

2. Harm and exploitation

You must not use the service to facilitate or meaningfully enable:

  • Violence, exploitation, trafficking, sexual abuse material or harm to children.
  • Harassment, hateful abuse, non-consensual intimate material or targeted intimidation.
  • Fraud, phishing, impersonation, deceptive manipulation or unlawful surveillance.
  • Malware, credential theft, unauthorised access or destructive cyber activity.

3. High-impact decisions

Do not use model output as the sole basis for decisions that determine legal rights, healthcare, employment, housing, insurance, credit, essential services or other significant interests. Use qualified human review, appropriate evidence and a lawful appeal path.

4. Security and platform integrity

Do not bypass authentication, policy, budget, quota, provider eligibility or safety controls. Do not probe, scan, load test, reverse engineer or attempt to disrupt the service without prior written authorisation.

5. Provider and route restrictions

You must comply with restrictions that apply to eligible model providers and routes. A policy alias does not remove provider-specific restrictions. Do not intentionally route prohibited workloads to a provider that does not permit them.

6. Tools and downstream actions

Treat model-generated tool arguments as untrusted. Validate inputs, verify user authority, apply least privilege, require approval for consequential actions and make irreversible operations appropriately difficult.

7. Enforcement

We may investigate suspected violations and limit, suspend or terminate affected access. Where practicable, we consider severity, intent, repetition, customer remediation and risk to people or systems.

8. Reporting

Report suspected abuse to abuse@octopusos.ai. Report security incidents affecting your credentials through the support path in your agreement or support@octopusos.ai.