SendRaven

Acceptable Use Policy

Last Updated: 2 September 2026

This policy says what you may send through SendRaven and what happens when you don't follow it. It is part of the Terms of Service. "You" means the Customer and everyone — and everything — that holds your API keys.

SendRaven sends through shared infrastructure. One customer's bad list raises the complaint rate that mailbox providers and our delivery provider measure for everyone on it. Most of this policy exists to keep that from happening.

1. Consent

Send marketing mail only to people who opted in to receive it from you. "Opted in" means they gave you their address and asked for your mail, and you can show when and how.

Do not send to purchased, rented, scraped or appended lists. An address you did not collect yourself is not opted in, whatever the list was described as.

Send transactional mail only where a relationship already exists: the recipient bought something, created an account, asked for a password reset, or otherwise triggered the message themselves. A "transactional" label on a message that promotes something to a stranger is a marketing message without consent.

2. Unsubscribe

Every marketing message carries a working unsubscribe link, and an unsubscribe takes effect immediately — not after a processing period, not at the next campaign. SendRaven inserts the link, records the opt-out, and cancels that recipient's pending scheduled sends and automation enrolments automatically.

You may not circumvent this. That includes removing or hiding the link, re-importing suppressed addresses, moving them to a different workspace or sending domain, or asking recipients to log in before they can opt out.

A marketing unsubscribe does not block transactional mail to the same address. A hard bounce or a spam complaint does.

3. Bounce and complaint thresholds

We monitor bounce and complaint rates for every sending identity. Our delivery provider places accounts under review at a bounce rate of about 5% and a complaint rate of about 0.1%, and at that point every customer on the platform is affected. We act before then.

We may refuse a campaign before it starts when the audience or the sending identity's recent history makes it likely to trip those rates. We may pause a sending identity, or a whole workspace, when its rates approach those levels, and keep it paused until we understand what happened. A pause is not a judgement that you did something wrong; a stale list produces the same numbers as a bad one.

Addresses that hard-bounce are suppressed automatically. Transient bounces (full mailbox, greylisting) are not; they resolve on their own.

4. Prohibited content and practices

You may not use SendRaven for:

  • Phishing, or any message designed to obtain passwords, payment details or other private information under false pretences.
  • Impersonation of any person or organisation, including spoofed sender names or addresses, and sending from a domain you do not own or are not authorised to use.
  • Malware, or links to it.
  • Mail that violates CAN-SPAM, GDPR, CASL, or any similar law that applies where you or your recipients are.
  • Content that is illegal in the recipient's jurisdiction.
  • Sending on behalf of third parties without disclosing it. Do not resell SendRaven, and do not run other companies' mail through your workspace, without a written agreement with us.

The general content restrictions in the Terms of Service apply as well.

5. Agents and automated senders

Sends made by autonomous software or AI agents holding your API keys are your sends. You are responsible for them exactly as if a person on your team had pressed the button, whether or not you reviewed the message first.

SendRaven gives every API key its own controls: a daily send limit, a recipient allowlist, and an approval hold that queues a message for a person to release. They exist for this case. Put them on every credential you hand to an agent, and size them to what the agent is actually supposed to do. An agent that is meant to answer support replies does not need to send to ten thousand addresses a day, and a key that lets it do so is a key that will, one day.

6. Inbound mail

Inbound processing is for reading and answering mail sent to your domains. You may not use it to harvest addresses, to build lists from messages you receive, or to relay mail — accepting a message on one side and sending it out the other, with or without changes, to recipients who did not write to you.

7. Enforcement

When something trips this policy we can refuse individual sends, hold a campaign, pause a sending identity or workspace, suspend your account, or terminate it. Which one depends on what happened and whether it is still happening.

We will tell you what tripped — the rate, the campaign, the credential — unless doing so would help an attacker. If your key was stolen and used for phishing, the details go to you after the key is revoked, not before.

Report abuse of SendRaven, including mail you received that you believe violates this policy, to abuse@sendraven.ai. Include the full message headers if you can; they tell us which sending identity to look at.

We may update this policy. Material changes are announced in the dashboard or by email, and apply from the date they are posted.