AI Services Terms

AI output may be inaccurate, incomplete, delayed, offensive, unexpected, or unsuitable for a particular purpose. You are responsible for reviewing, testing, supervising, and approving how AI features are configured and used.

Elav8 AI LLC does not warrant that AI output is accurate, lawful for your specific use case, non-infringing, appropriate for every recipient, or free from hallucinations or bias.

Similar inputs may not always produce similar outputs. AI features can misunderstand context, omit information, fabricate facts, overstate certainty, or fail to follow instructions. Human review and guardrails are required for high-risk or sensitive workflows.

AI output is not legal, medical, financial, tax, employment, insurance, emergency, or other professional advice. You are responsible for obtaining qualified professional review where required.

You are responsible for prompts, scripts, objectives, knowledge sources, call flows, message templates, escalation rules, disclosures, consent, opt-outs, and any action taken based on AI output.

You must not use AI features to make unlawful decisions, deceive recipients, impersonate people, collect sensitive information without authorization, or provide regulated advice unless you have appropriate legal authority and written approval from Provider.

You must test agents before activation, monitor live performance, review transcripts and summaries, maintain accurate knowledge sources, configure escalation paths, and promptly correct or disable workflows that produce unsafe, inaccurate, unwanted, or non-compliant behavior.

You are responsible for ensuring any required AI, automated-call, call-recording, telemarketing, consumer-protection, and privacy disclosures are made at the right time and in the right form for your use case and jurisdictions.

AI features may send inputs and context to model providers and infrastructure subprocessors as needed to provide the Service. Inputs may include prompts, scripts, knowledge snippets, conversation history, call audio, transcripts, messages, metadata, and configuration.

We use commercially reasonable controls to limit use of Customer Data to providing, securing, supporting, and improving the Service, subject to the Agreement and Privacy Policy.

Model providers, voice providers, transcription providers, and other AI subprocessors may process inputs and outputs according to agreements between those providers and Provider. AI features may change as providers update models, safety systems, latency, pricing, limits, or API behavior.

Provider may log prompts, outputs, traces, tool calls, evaluation results, and safety signals to debug, monitor, secure, improve, and document the Service, subject to the Privacy Policy and applicable customer settings.

You should test agents before activation and monitor performance after activation. The Service may provide guardrails, logs, transcripts, summaries, and escalation features, but those features do not remove your responsibility for lawful and appropriate use.

Elav8 AI LLC may suspend or limit AI features if configuration or behavior creates compliance, safety, carrier, reputational, security, or operational risk.

Provider may run automated or manual evaluations for quality, safety, abuse, policy compliance, deliverability, and security. These checks are advisory and do not guarantee lawful or error-free behavior.

You must not disable, bypass, or misconfigure safeguards in a way that increases risk to recipients, customers, carriers, Provider, or third parties.

Unless a signed agreement says otherwise, Provider does not promise that any AI model will or will not train on particular data beyond the commitments in its applicable subprocessors and Privacy Policy. Provider may use aggregated, de-identified, or synthetic information to improve reliability, safety, analytics, and product performance.

AI features, voices, model versions, latency, limits, supported languages, safety behavior, and pricing may change over time. Provider may substitute equivalent or improved model providers or features where reasonable.

AI features must not be treated as a substitute for human authorization where a human signature, payment authorization, legal consent, compliance certification, or regulated approval is required. In the registration flow, consent checkboxes, signatures, and payment actions must be completed by an authorized human.

Customer remains responsible for commitments, representations, discounts, appointments, bookings, purchases, or operational instructions made or accepted through its configured AI workflows.