1. Prohibited activity
- Illegal, fraudulent, deceptive, threatening, discriminatory or harmful activity.
- Spam, unlawful direct marketing or communication without an appropriate legal basis.
- Impersonating people or concealing that a user is interacting with AI where disclosure is required.
- Collecting passwords, payment-card data, health data or other sensitive information without an approved design and agreement.
- Malware, security testing without permission, credential abuse or attempts to bypass usage controls.
- Using the service to make high-impact legal, medical, employment, insurance, credit or eligibility decisions.
2. Messaging and calls
Customers and partners must have the required consent or other legal authority for contact, honour opt-outs, identify the sender, respect applicable calling and messaging restrictions, and configure recording disclosures where calls are recorded or transcribed.
3. AI use
AI output must be treated as assistive, not guaranteed fact. Customers must provide accurate source information, test workflows, define escalation rules and retain qualified human oversight for sensitive or consequential interactions.
4. Partner responsibility
Partners may not make unsupported performance claims, hide required disclosures, sell unrestricted access, misrepresent BuiltBehind capabilities or deploy unapproved high-risk workflows. White-label access remains conditional on operational and compliance standards.
5. Enforcement
The final service agreement should permit investigation, rate limiting, workflow suspension or termination where use threatens people, systems, legal compliance or platform integrity, with proportionate notice where appropriate.