Lumoswitch Acceptable Use Policy

1. Lawful and authorized use

Use Lumoswitch only where lawful, properly authorized, and consistent with the policies of your chosen model provider. You are responsible for inputs, model selection, use of output, and final business outcomes.

2. Prohibited content and conduct

• Committing, assisting, or encouraging crime, terrorism, violent harm, human trafficking, or trade in controlled goods.
• Generating or distributing child sexual abuse material, or exploiting, grooming, or harming minors in any way.
• Infringing privacy, personal information, trade secrets, intellectual property, reputation, or other lawful rights.
• Phishing, fraud, impersonation, spam, malicious marketing, fake traffic, or manipulative conduct.
• Developing or distributing malicious code; attacking, probing, disrupting, or bypassing system access controls or security protections.
• Unauthorized collection of personal information, account cracking, credential stuffing, evading rate limits, or reselling accounts and access keys.
• Using model output without professional review to directly decide matters affecting healthcare, credit, employment, insurance, justice, or other significant individual rights.
• Violating model providers' use policies, sanctions, or export-control requirements, or using routing to evade third-party restrictions.

3. High-risk uses

For healthcare, law, finance, critical infrastructure, public safety, and similar uses, you must provide human review, access controls, audit logs, and emergency procedures, and clearly inform affected people of AI involvement. Lumoswitch does not provide professional advice in these fields.

4. Security and fair use

• Lumoswitch inference services may enforce authentication, model authorization, budgets, rate limits, and runtime security policies. Lumoswitch may automatically check, block, or manually review public community content. Actual capabilities depend on the features enabled in the current deployment.
• Follow API quotas, concurrency and rate limits, and reasonable-use requirements.
• Do not disrupt other users, exhaust shared resources, or impose unreasonable load on the platform.
• Report vulnerabilities responsibly. Do not exploit them to access, change, delete, or disclose data.

5. Enforcement

Depending on severity, we may warn, rate-limit, block requests, revoke keys, suspend accounts, restrict sign-in, or terminate service. For clearly unlawful activity, real risks to physical safety, or regulatory requirements, we may preserve necessary evidence and cooperate with competent authorities as required by law.

If you believe an action was mistaken, appeal to soraincloud@hakihakii.com with the necessary explanation.