Model Use Agreement
Last Updated: August 2026
1. AI Model Characteristics & Non-Determinism
Our foundation models, speech engines, and video generation pipelines produce probabilistic completions. Outputs may vary across identical prompts and may occasionally contain inaccuracies. You are solely responsible for verifying the accuracy, safety, and regulatory compliance of all Output before relying on it for high-stakes operational, financial, or educational applications.
2. Voice Cloning & Biometric Synthesis Ethics
Our speech platform provides pre-calibrated dual-host voice personas. When using custom voice cloning features, you must possess explicit, verifiable written consent from the voice owner. You agree not to: (a) create deceptive voice impersonations; (b) clone individuals without authorization; or (c) use synthetic voices to misrepresent identity in telephone communications.
3. Autonomous Agent Permissions & Human Oversight
Autonomous agents, such as those used for course publishing, order intake, and conversational messaging, execute actions based on structured tool schemas. Organizations must implement appropriate human-in-the-loop oversight and threshold validations for high-value financial actions or critical database modifications.
4. Content Moderation, Rate Limiting & Enforcement
We deploy automated safety classifiers, input guardrails, and rate-limiting monitors to detect abusive, hateful, or hazardous prompts. We reserve the right to throttle, restrict, or terminate API access for accounts that repeatedly violate these Model Use standards.
5. High-Risk Use Disclaimers
Our models and autonomous agents are not certified for use in critical life-support systems, emergency response dispatch, direct medical diagnosis, or automated weapons control.