AI Usage Policy

The Service uses large language models for text generation, classification, and analysis, and voice models for the Talk voice agent. Models are provided by third-party providers under contracts that restrict how they may handle your data.

Business communication with your customers, customer support, drafting content and social posts, summarizing calls and messages, and analyzing your own business data.

You may not use the Service to generate misinformation, to impersonate people or organizations deceptively, to manipulate political processes, to harass, or to create content that facilitates harm or illegal activity.

The Acceptable Content Policy lists prohibited content in detail.

You are accountable for what your business sends. Review AI output before relying on it, correct errors, and configure your agents so they represent your business accurately.

AI output can contain errors, fabrications, or biased content. It is not professional advice. Do not treat AI output as legal, medical, or financial advice.

Content sent to AI model providers is used to generate a response and is not used to train their models. See the AI Data Training Policy for the full commitment.

The platform provides review and approval steps for consequential actions such as publishing content or sending campaigns. We recommend keeping a human in the loop for anything customer-facing that is new or sensitive.

Where the law requires that automated or AI communication be disclosed, you must configure your use accordingly. Fulcrum voice and chat agents support disclosure that a caller or visitor is speaking with an AI assistant.

You must comply with AI-related laws that apply to you, including the EU AI Act where relevant, applicable US federal and state law, and the Australian AI Ethics Principles as reflected in our AI Ethics Framework.