AI Terms

Effective November 12, 2026

These AI Terms are part of the Prevision Terms of Service. They apply when you use Lexi or any other feature of the Service that uses artificial intelligence ("AI Features"), including AI texting, AI calling, call summaries and the assistant.

1. What AI Features do

AI Features generate text, speech and summaries with models from third-party providers. They act only within the settings you choose, such as which leads Lexi works, your hours, your booking rules and your script.

2. Output can be wrong

AI output is produced by a statistical model. It can be inaccurate, incomplete, out of date, or inappropriate, and it can sound confident when it is wrong. Do not rely on AI output as the only basis for any decision.

3. You supervise, and the output is yours

You are responsible for turning AI Features on and off, for their settings, for monitoring their conversations, and for every message, call and statement they make on your behalf. Review conversations regularly and step in when needed.

4. Not a licensed producer

AI Features are not licensed insurance producers. They must not give insurance advice, recommend a product, quote a binding price, make a coverage or underwriting decision, tell a person whether they qualify, or take an application. They may answer simple questions, gather information you ask for, and book appointments with you or a licensed producer. You are responsible for configuring them to stay within those limits and for anything they say beyond them.

5. Disclosure

AI Features are set to say they are automated when a person asks. Some laws require disclosure in other cases, such as at the start of an AI voice call or when AI is used to encourage a sale. You are responsible for meeting any disclosure rule that applies to you, and the Messaging and Calling Terms apply to every AI text and call.

6. Your data and the models

To produce output, the text needed for the task is sent to our model provider. We choose AI providers whose terms bar training their models on this data. Health answers stored in a client's intake are never sent to an AI model. Health details a person mentions in a text or a call may be processed, for example in a call summary. Do not put information into an AI Feature that you are not allowed to share with a service provider.

7. Our rights

We may change the models and providers behind AI Features, limit their use, or turn an AI Feature off for an account if we believe it is being misused or causing harm.

8. No warranty

AI Features are provided as is, as section 16 of the Terms of Service describes. Our liability for AI Features is limited as section 17 of the Terms of Service provides.