Last updated: September 26, 2026

AI Disclosure

This document is provided in English only. The English version is the controlling text.

1. Overview

This AI Disclosure explains the artificial intelligence features available in embedio, what data they use, and the controls available to you. It is part of, and incorporated into, our Terms of Service and Privacy Policy.

2. AI Chat

AI Chat is an optional, paid module a customer can enable for their project. When enabled, it answers an End User's questions in a chat widget embedded on the customer's site, using a large language model operated by a third-party AI provider together with the FAQ and knowledge-base content the customer has configured.

If AI Chat cannot resolve a question, or an End User asks to speak with a person, the conversation can be escalated to the customer's own Live Chat module, if enabled, where a human on the customer's team responds directly.

3. What Is Not Generative AI

embedio also uses automated, rules-based scoring, not a generative AI model, to flag potentially spammy or abusive form submissions, such as unusually rapid submission bursts or repeated identical content. This is a heuristic signal for the customer to review, not a fully automated decision; the customer can always unflag a submission.

4. What Data Is Sent to the AI Provider

When an End User sends a message in AI Chat, we send that message, recent conversation history, and the customer's configured knowledge base and FAQ entries to our AI provider's API to generate a reply. We do not send the customer's other, unrelated account data, billing information, or other End Users' data.

Our AI provider may process this data on servers located outside the United States. We select AI providers that contractually commit to appropriate data protection and security obligations, but we do not control, and this Policy does not override, that provider's own infrastructure or data-handling practices.

5. Retention and Model Training

Conversation transcripts are stored in our database and are visible to the customer who owns the project, for their own record-keeping and quality review, and to our own support staff when needed to help troubleshoot an issue you report. We do not use this conversation data to train our own models. Our AI provider processes the request to generate the response under its own data processing terms; we do not control, and this Policy does not override, that provider's own data retention practices for API traffic, which we encourage you to review if you have specific concerns.

6. Accuracy and Limitations

AI-generated responses are produced by a statistical model and may be incomplete, out of date, or simply wrong, including about a customer's own products, policies, or availability. AI Chat is not a substitute for professional advice, and neither embedio nor the underlying model is responsible for decisions made in reliance on its output. Customers are responsible for reviewing their knowledge base content and monitoring conversations for accuracy.

7. Your Controls

AI Chat is opt-in: a customer can enable or disable it at any time from Billing, and can edit or remove knowledge-base content at any time. Once disabled, no further End-User messages for that project are sent to the AI provider. Customers can review full conversation history for their own project in the AI Chat section of their dashboard.

8. Automated Decision-Making

AI Chat is designed to answer questions and route conversations to a human, not to make legal or similarly significant decisions about individuals; for example, it does not approve or deny applications, pricing, or eligibility. If a customer configures a use case that does this, the customer, not embedio, is responsible for ensuring appropriate human oversight and compliance with applicable law, consistent with our Acceptable Use Policy.

9. Contact

Questions about our AI features can be sent to contact@tryembedio.com, or through the Support option in your dashboard.