Acceptable Use Policy

Replicas is a powerful tool. This policy explains what responsible use looks like and where we draw the line.

• Use Replicas for lawful market research, product validation, audience analysis, and business decision-making.
• Do not use Replicas to make or inform consequential decisions about real individuals, including hiring, lending, insurance underwriting, housing, or legal proceedings.
• Do not use replica outputs to impersonate real, identifiable individuals or to create fake testimonials attributed to real people.
• Do not present replica outputs as real human opinions. Always disclose that research was conducted with synthetic respondents when sharing results externally.
• Do not use Replicas to generate hateful, harassing, sexually explicit, or violent content targeting any individual or group.
• Do not use Replicas to produce spam, phishing material, malware, or any form of automated abuse.
• Do not attempt to reverse-engineer, extract, or reconstruct the underlying language models, training data, or proprietary algorithms used by Replicas.
• Do not scrape, bulk-export, or systematically extract replica outputs for the purpose of training other AI models.
• Do not intentionally abuse API rate limits, overload infrastructure, or interfere with the service for other users.
• Do not use Replicas to circumvent laws or regulations in your jurisdiction, including consumer protection, advertising standards, and data protection laws.
• Violations of this policy may result in suspension or permanent termination of your account. We will attempt to notify you before taking action except in cases of severe or repeated abuse.
• If you are unsure whether a use case is acceptable, contact us before proceeding. We are happy to help.