1. You are responsible for your Twin
Every action your Twin executes — autonomous or supervised — is attributable to you. Configure autonomy levels and approval rules accordingly.
2. Prohibited content
- Sexual content involving minors, or any content that sexualizes real people without consent.
- Content that incites violence, self-harm, terrorism, or genocide.
- Targeted harassment, doxxing, or coordinated abuse.
- Disinformation campaigns, deepfakes designed to deceive, or impersonation of real people without their explicit consent.
- Material that violates intellectual-property rights or trade-secret obligations.
3. Prohibited activity
- Reverse engineering, scraping outputs at scale, or training competing foundation models on Lumen Twin outputs.
- Bypassing rate limits, quotas, or safety classifiers.
- Using the API to send unsolicited mass communications (spam).
- Operating critical-safety systems (medical diagnostics, legal final advice, autonomous vehicles, life-support) without an independent human authority of control.
4. Voice cloning
You may only clone voices you own or for which you hold explicit, revocable, written consent. We sample-watermark synthesized audio and reserve the right to suspend voice features at any time.
5. Autonomous actions
When your Twin acts autonomously (sends emails, books meetings, transacts), you remain the legally responsible party. We provide audit trails — keep them.
6. Enforcement
We may rate-limit, suspend, or terminate accounts that violate this policy. Severe violations (CSAM, coordinated harm) are reported to the relevant authorities.
7. Reporting
Suspected abuse: support@lumendralabs.com. Security issues: support@lumendralabs.com.