AI Usage Policy

This page explains how the AI features of outr work, what data they use, where they fall short, and what you are responsible for when you use them. It is part of our Terms of Service, and the same defined words apply here.

outr is an AI agent (called Mira) that builds and runs B2B cold email. This policy covers whenever you use those AI features.

1. What the AI does

The AI helps you prepare and run B2B outreach. It can:

• draft email sequences from the business profile and targeting you give it;
• draft follow-ups for a sequence;
• draft suggested replies to messages prospects send back; and
• answer questions about your setup, your plan, and how to use outr.

The AI produces drafts and suggestions. It is a drafting and assistance tool that works from the information and instructions you give it. It does not make outreach decisions for you.

2. Nothing sends without your approval

The AI never sends outreach on its own. Everything it produces is a draft you review. A campaign starts as a draft, and no email goes to any recipient until you review the campaign and click Approve and launch. No background job or automation launches a campaign for you. You decide who is contacted, and you approve the content and the list before anything sends, as set out in the Outreach Model section of the Terms of Service.

3. Which AI we use

The AI features run on a large language model from a third-party AI provider, which we reach through an AI gateway. We may use different model tiers for different tasks, and we may change or add AI providers or models over time to run and improve outr. If we do, we will keep this policy accurate.

4. What data goes to the AI, and what does not

To produce useful drafts, outr sends the following to the AI as the task needs:

• your business and campaign context, such as your company, offer, value propositions, targeting (role, industry, region, company size), tone, and signature details;
• the task context for what you asked the AI to do; and
• when you ask the AI to draft a reply, the text of the prospect's incoming message, so the draft can respond to what they actually wrote.

We do not send your payment card details to the AI. Card data is handled entirely by our payment processor and is never used by the AI. The AI receives only what the task needs.

The AI provider is one of our sub-processors. The data it handles and its role are described in our Privacy Policy and, where it applies, our Data Processing Addendum.

5. Where AI output falls short

Treat AI output as a starting draft, not a finished or verified result. In particular:

• It can be wrong. It may be inaccurate, incomplete, outdated, biased, or unsuitable for your purpose, and it can occasionally state things that are not true.
• It is not legal advice. It is not legal, tax, compliance, or other professional advice, and you must not rely on anything the AI says about compliance as if it were.
• It is not a promise of results. It does not guarantee replies, meetings, leads, deliverability, or inbox placement.
• You review it before you use it. You are responsible for checking every AI draft for accuracy, tone, legality, and fit before you approve, send, or rely on it.

outr runs automated checks on generated email copy (for example, formatting and content checks and a fallback to a safe template when a draft looks unsafe or malformed), and it will decline to draft in some situations rather than produce low-quality or unsuitable output. These checks reduce the risk of unsuitable output but do not remove it, and they do not replace your review.

6. You are responsible for what you approve

You are responsible for any content you approve, send, or act on, including content that started as an AI draft. Once you approve a draft, it is your content and your outreach, and the responsibilities in the Outreach Model section and the Acceptable Use Policy apply to it in full, including your promises about lawful basis and consent, and your responsibility for the legality of your outreach.

7. What you may not use the AI for

You may not use the AI features to:

• generate content that would break the Acceptable Use Policy if you sent it, including unlawful, deceptive, harassing, phishing, fraudulent, adult, or infringing content;
• get around or defeat the guardrails, the approval gate, or the safety checks;
• generate content you know or should know is false, misleading, or meant to deceive a recipient; or
• use the AI for anything outside legitimate B2B outreach as allowed by the Terms of Service.

outr includes automated guardrails that refuse certain requests (for example, consumer or purchased-list requests, harmful content, out-of-scope channels, and abuse) before any AI processing happens. These guardrails support this policy but do not replace your own responsibility to use the AI lawfully and in good faith.

8. Training on your data

outr does not use your customer data or prospect reply text to train its own models. We do not build or train AI models on your content.

The AI features are run by our AI provider on our behalf. We will only say that the AI provider does not train on your content and prospect reply text once that is confirmed in our agreement with them. We rely on that agreement for this and will keep this policy accurate as those terms are confirmed and updated. If you have a specific question about how the AI provider handles data, contact us and we will share what we can.

Any use of aggregated, de-identified data to run and improve outr is limited as set out in the Terms of Service and does not identify you or any recipient.

9. Changes

We may update this policy, including to reflect a change in the AI providers or models we use. The date at the top shows when the current version started. For a material change we will give notice as set out in the Terms of Service. Using the AI features after a change takes effect means you accept the updated policy.

10. Contact

Questions about this policy, about how the AI features handle data, or any privacy question: support@tryoutr.io.