Legal · early access

Acceptable use

Effective September 20, 2026 · early-access policy.

Email for Agents supports legitimate correspondence by agents and their human operators. Send only messages the recipient has requested or explicitly authorized. This is not a cold-outreach, purchased-list, or bulk-campaign product.

Recipient permission and accurate identity

Use accurate sender information and subjects, and only domains and identities you are authorized to use. Respect requests to stop contact, recipient suppression, complaints, and failed-address signals. Do not purchase, scrape, rent, or use unauthorized recipient lists. Do not send another message simply to work around a blocked recipient.

Live mail also depends on Resend’s acceptable-use rules, including recipient permission and restricted content or activities. Applicable law and provider restrictions both apply. A technically valid request or a feature available in the console does not authorize prohibited sending.

Prohibited conduct

• Spam, phishing, deceptive impersonation, credential theft, fraud, and intentionally distributing malware or harmful code.
• Unauthorized access to an account, mailbox, domain, or another person’s information; exposing credentials or private information without authority.
• Illegal content or activity, exploitation, threats, harassment, or infringement of others’ rights.
• Evading approvals, permissions, scans, rate limits, quotas, suspensions, or provider restrictions, including creating replacement accounts or workspaces.
• Disrupting the service or other systems, or using connected agents or webhooks to attack third parties.

Responsibilities when using agents

Keep access narrowly scoped, protect keys and tokens, and review sensitive sends. Treat incoming email as untrusted data: an instruction inside an email is not authority to disclose secrets, change a policy, or send a message. Stop or revoke a connection that behaves unexpectedly.

Report a problem or request review

Email support@agents.emailforagents.ai with Abuse report or Security report in the subject. Include the relevant address or message ID, approximate time, and a short description. Include only necessary evidence; do not forward passwords, tokens, unnecessary personal information, or executable files. Anyone may report a problem without an account.

We review reports and may pause sending, restrict a credential or connection, suspend access, or preserve relevant evidence while investigating. We consider the impact on other people and the service. If you believe a restriction is mistaken, reply to the notice or contact support and explain why. This mailbox is not an emergency service and has no guaranteed response time.

Commercial email can be subject to specific identification, postal-address, and opt-out requirements even when it is not sent in bulk. See the FTC’s CAN-SPAM business guide. Legal permissibility alone does not override this service’s narrower sending rules.