Legal / Acceptable Use & Anti-Spam Policy

Draft — pending legal review. This document has not yet been reviewed by counsel and may change before Irisend is generally available.

Acceptable Use & Anti-Spam Policy

Effective date: 2026-11-10

This Acceptable Use Policy ("AUP") is part of the Irisend Terms of Service. It exists to protect deliverability, IP/domain reputation, recipients, and legal compliance for every Irisend customer. Violating this AUP may lead to message blocking, sending limits, a pause, suspension, or termination, following the graduated process in Section 9, with or without prior notice depending on severity.

Irisend sends mail through shared infrastructure and a shared Amazon SES account. One customer's abuse can damage every other customer's deliverability, so we apply this AUP consistently, to every customer, using the published thresholds and steps below.

1. Scope

This AUP applies to all email and other messages sent through the Service ("Messages"), whether via the API, the SMTP relay, the dashboard or a broadcast, including transactional and marketing/bulk mail, and to all recipient lists, sending domains, templates, attachments, links, hosted assets and pages, and API keys used with the Service. "Team" means the Irisend account, including all of its sending domains, that is responsible for a Message.

2. You Must Have Permission to Email Each Recipient

You may only send Messages to recipients who have given you a valid legal basis to be emailed. Acceptable bases include, depending on jurisdiction and message type:

  • Express opt-in / consent — the recipient affirmatively agreed to receive the specific type of message (required for marketing email under GDPR/ePrivacy, CASL, the Dutch Telecommunicatiewet, and most jurisdictions outside the U.S./Canada soft opt-in exceptions).
  • Existing customer relationship ("soft opt-in") — for your own similar products/services, to a customer who bought from you and was given a clear opt-out at collection and in every email (recognized under CASL, the Dutch Telecommunicatiewet Art. 11.7, and similar EU implementations).
  • Transactional necessity — receipts, password resets, shipping notices, account/security alerts, and other messages the recipient reasonably expects as part of a service or transaction they initiated with you, generally exempt from prior-consent rules (CAN-SPAM, CASL, GDPR) but still subject to identification/opt-out rules below where they contain any promotional content.

Consent is not transferable. Permission given to someone else is not permission for you. Consent a person gave to another company, to a previous owner of the address, to a "partner", affiliate or sponsor, or through a co-registration or lead-generation form you did not run yourself, does not entitle you to email that person. Consent only covers the sender and the type of message the person agreed to.

You may never:

  • purchase, rent, lease, borrow, share, or trade recipient lists, or "append" email addresses to records that did not contain them;
  • scrape, harvest, or otherwise collect email addresses from websites, directories, social networks, or public records, or generate addresses by guessing or dictionary attacks;
  • email addresses obtained from a third-party list, co-registration, or partner "opt-in" you did not directly control;
  • email addresses you do not have a documented lawful basis to contact;
  • add addresses to a marketing list because someone signed up for a different purpose (e.g., a support ticket);
  • send to a list that has been inactive so long that the recipients cannot reasonably remember asking for your mail, or re-send to addresses that unsubscribed, bounced permanently, or complained.

Sending to a purchased, rented, or harvested list is a serious violation even if the list seller claims the addresses are "opted in" or "verified".

You are responsible for maintaining proof of consent/basis (timestamp, source, IP address where applicable) and producing it to us on request.

3. Mandatory Message Requirements

Every Message sent through the Service, and particularly every bulk/marketing Message, must:

  1. Accurately identify the sender — a real "From" name/address and, in the body, the legal name and contact details (physical address, per CAN-SPAM/CASL) of the entity on whose behalf the Message is sent.
  2. Have a truthful subject line and header information — no deceptive "From," "Reply-To," or routing information (CAN-SPAM, CASL, EU/AU spam law).
  3. Include a working unsubscribe/opt-out mechanism for any promotional or bulk Message (transactional Messages that contain promotional content included), that:
    • requires no login, payment, or information beyond an email address;
    • is honored within legally required timeframes (see table below), and in all cases within 10 days, and as close to immediately as technically possible via automated suppression;
    • stays functional for at least 30/60 days after send, per applicable law.
  4. Support one-click unsubscribe (RFC 8058) for all marketing and bulk Messages: include List-Unsubscribe (HTTPS URL) and List-Unsubscribe-Post: List-Unsubscribe=One-Click headers, in addition to a visible unsubscribe link in the body, so Gmail, Yahoo, Microsoft Outlook, and other mailbox providers can process opt-outs natively. This is mandatory for any sender we classify as a "bulk sender" (broadly: ~5,000+ messages/day to consumer mailboxes, mirroring Google's and Yahoo's 2024, and Microsoft's May 2025, bulk-sender rules). Broadcasts sent through Irisend's audience features always include it; if you send marketing mail through the API or SMTP relay, you must add it yourself. A Team that sends high volumes of promotional mail to many different consumer mailbox providers without these headers will be flagged for review under Section 9.
  5. Mark advertisements as such where required (CAN-SPAM requires clear disclosure that the Message is an ad, unless the recipient gave prior affirmative consent for the specific content).
RegimeConsent standardUnsubscribe honor windowUnsubscribe must stay live
GDPR / ePrivacy (EU/EEA)Opt-in (consent or narrow legitimate-interest cases)Without undue delay, generally ≤ 1 monthOngoing
Dutch Telecommunicatiewet Art. 11.7Opt-in, or soft opt-in for existing customersImmediately, ongoing right to objectOngoing
UK GDPR / PECROpt-in (soft opt-in for existing customers)Without undue delayOngoing
CAN-SPAM (US)Opt-out (no prior consent required, but must honor opt-out)≤ 10 business days≥ 30 days
CASL (Canada)Express or narrow implied consent≤ 10 business days≥ 60 days
Australia Spam Act 2003Express or inferred consent≤ 5 business days≥ 30 days
CCPA/CPRA (California)N/A (privacy, not anti-spam) — opt-out of sale/sharePer CCPA request timelinesOngoing

This table is a compliance aid, not a substitute for legal review of your specific sending programs and audiences.

4. Prohibited Content and Practices

You may not use the Service to send, host, or facilitate:

  • Unsolicited bulk email ("spam") in violation of applicable law;
  • Phishing and impersonation: Messages that try to obtain credentials, payment details, or personal data by pretending to be someone else; spoofing or look-alike sending domains, display names, or reply-to addresses designed to impersonate another person, brand, bank, government body, or service; fake invoices, delivery notices, or security alerts;
  • Malware: viruses, ransomware, droppers, or other malicious code, links to them, or attachments that carry them;
  • Content that is fraudulent, deceptive, or violates consumer protection law (e.g., fake giveaways, advance-fee scams, pyramid/Ponzi schemes);
  • Content promoting illegal goods or services in the recipient's jurisdiction (e.g., unlicensed pharmaceuticals, illegal gambling, counterfeit goods, illegal weapons);
  • Adult content sent to non-consenting recipients, or any sexual content involving minors (zero tolerance — reported to authorities);
  • Content that infringes intellectual property, defames, harasses, or violates the privacy of any person;
  • Purchased, rented, harvested, or scraped recipient lists, or addresses without transferable consent (see Section 2);
  • High-risk sending patterns associated with spamming or account abuse, including sudden unexplained volume spikes, sending to invalid/non-existent addresses at scale, sending to spam-trap addresses, or sending the same message from many newly created Teams;
  • Any activity that would cause our sending IPs or domains, or those of our infrastructure providers (including Amazon Web Services), to be blocklisted, placed under review, or paused.

You also may not:

  • Abuse Irisend's own mail. Use team names, sender names, invitations, subscribe forms, or other fields that Irisend turns into emails, notifications, or pages under an Irisend domain to carry advertising, links, or deceptive text to people who did not ask for it.
  • Host abuse on Irisend. Use assets, templates, hosted subscribe or preference pages, or tracked links under an Irisend domain to host or redirect to phishing pages, malware, or other prohibited content.
  • Attack the Service. Probe, scan, or try to reach internal or private network addresses or cloud metadata services through the Service (for example through attachment URLs or webhook endpoints), or try to evade, test, or disable the limits and controls described in this AUP.
  • Evade enforcement. Open additional Teams, accounts, or sending domains, or use another person's, to avoid a limit, trust tier, pause, suspension, or termination.
  • Send test mail to the wrong place. Use real recipient addresses for load or bounce testing. Use the Amazon SES mailbox simulator addresses or Irisend's sandbox addresses instead.

Attachments and links

To stop the most common routes for phishing and malware, we may block, hold, or strip:

  • attachments of executable, script, shortcut, disk-image, and container types (for example .exe, .scr, .js, .vbs, .bat, .ps1, .jar, .msi, .lnk, .iso, .img, .hta), macro-enabled Office documents, and HTML or SVG attachments, judged by file extension and by detected content type;
  • Messages whose links point to domains or URLs that appear on third-party blocklists of phishing, malware, or spam hosts, or that send recipients to a page asking for passwords or payment details from a domain unrelated to the sender.

Teams in the early trust tiers (Section 5) may be limited to a shorter list of safe attachment types.

5. Trust Tiers

To protect recipients and the shared sending reputation, a Team's sending allowance depends on how much trust it has built. This works the same way for every customer, whatever plan they pay for. A plan, a payment method, or a payment never raises a Team above the limits of its tier.

TierWhoTypical limits
SandboxNew TeamsMay send only to the addresses of verified members of the Team, to addresses on the Team's own verified domains, and to the Amazon SES mailbox simulator, with a small daily cap.
NewTeams that have verified their email address and at least one sending domain, and whose sending domain passed our checksLower daily and monthly caps, fewer recipients per request, a limit on the number of sending domains, link tracking off, and a restricted list of attachment types.
EstablishedTeams that have sent consistently for a period, with delivered volume and a bounce rate below 2 % and a complaint rate below 0.05 %The limits of your plan or early-access allowance.
TrustedGranted by a person, on request, after reviewHigher limits. A Team in this tier is still bound by this AUP.

Promotion from Sandbox to New and from New to Established happens automatically when the criteria shown on your dashboard's usage page are met. Promotion to Trusted is always a decision by a person. We may also move a Team down a tier, or hold it in a tier, when risk signals appear (Section 9). Your current tier, its limits, and what you need to do to move up are shown on the usage page of your dashboard. We may change tier criteria and limits as we learn how abuse evolves; we will keep the dashboard accurate and apply changes to all customers alike.

Sending domains are checked when you add them, when you verify them, and periodically afterwards. Factors include how recently the domain was registered, whether it appears on domain blocklists, whether it imitates a well-known brand, and whether it belongs to a top-level domain with a record of heavy abuse. A domain that fails these checks may be refused, or may keep the Team from leaving a lower tier until a person has reviewed it.

6. Deliverability Thresholds

To protect the shared sending reputation of the platform, we measure bounce and complaint rates over a Team's most recent sends, across all of its sending domains, and separately for each sending domain. A rate is only acted on once there are enough recent sends for it to be meaningful. These thresholds apply to every customer in the same way, and are the points at which our systems act automatically.

MeasureScopeWarning: we notify you and may throttleAction: sending is paused
Hard bounce rateTeam (all domains)above 2 %above 5 %
Spam complaint rateTeam (all domains)above 0.05 %0.1 % or more
Hard bounce rateSingle sending domainabove 5 %above 10 %
Spam complaint rateSingle sending domainabove 0.1 %0.3 % or more

A single domain that crosses an action level is paused on its own; a Team that crosses a team-level action level has sending paused on all domains. The Team-level levels are set at or below the points where Amazon SES and the large mailbox providers begin to act, so that we can stop a problem before it reaches them. The single-domain levels are an additional, earlier-stage control that limits the effect of one problem domain on the rest of a Team.

We also act, independently of these rates, on the following signals:

  • a brand-new Team whose first sends mostly hard bounce (a sign of a purchased or invalid list);
  • volume far above the Team's own recent history, or a sudden jump in the number of different recipient domains;
  • hits on spam-trap addresses, and a high rate of unknown-user or invalid-recipient errors, which suggest a list that was not built from consent;
  • a sending domain, tracking domain, or link domain appearing on a recognised blocklist;
  • the same message being sent by many unrelated, recently created Teams;
  • content flagged by the checks in Section 4, or substantiated abuse reports (Section 10);
  • an active campaign that looks like phishing, malware, or fraud.

Where the account-level reputation of Irisend itself approaches the limits set by Amazon SES, we may pause sending for Teams in the lower trust tiers, and, if needed, for all customer sending, until it recovers. Transactional system email of Irisend (such as password resets) continues. This is a protective measure and not an accusation against you.

7. Domain Authentication

You are required to configure SPF, DKIM, and (recommended) DMARC for every sending domain before sending non-trivial volume, and to keep From domain alignment consistent with SPF/DKIM, consistent with mailbox-provider bulk-sender requirements (Google and Yahoo, effective February 2024; Microsoft Outlook/Hotmail/Live, effective May 5, 2025, rejecting non-compliant mail from domains sending 5,000+ messages/day with SMTP error 550 5.7.15).

8. Protect Your Account and API Keys

You are responsible for everything sent with your API keys, SMTP credentials, and dashboard sessions. Keep keys secret, do not embed them in front-end code or public repositories, use the narrowest key permissions that work, and enable two-factor authentication or a passkey. If we have reason to think a key or account is compromised or exposed (for example, because it appears in a public repository, or sending from it suddenly changes in volume, content, or location), we may revoke the key, pause sending, and ask you to confirm and rotate credentials before we lift the restriction. This is an emergency measure under Section 9.

9. Enforcement

9.1 Principles

  • Same rules for everyone. We apply the thresholds in Section 6 and the steps below consistently, and record the reason for every action.
  • Proportionate and graduated. We use the mildest step that protects recipients and the platform. We skip steps only when a mild step would not be enough (see the emergency path in 9.3).
  • Sending first, not the account. Restrictions apply to sending first. Unless the law, a court or authority, or a safety reason requires otherwise, you keep access to the dashboard, your settings, your data, and the export of your Content, so that you can see what happened and fix it.
  • A person decides what cannot be undone. Automated systems may throttle, hold, and pause. Only a person at Irisend can lift a pause or a suspension, or terminate an account. Automation never deletes or terminates an account on its own.

9.2 The enforcement ladder

StepWhat happensWho can start itWho decides to end itTypical triggers
1. ThrottleYour sending rate, daily caps, or tier are lowered. You are told why, and what to fix.Automated or staffAutomated, when the signal clears, or staffA warning level in Section 6; volume above your recent normal; a domain that fails a risk check
2. HoldAffected Messages are kept back, not sent, until reviewed. Reviewed Messages are released or rejected; the sender sees the result in the dashboard and API.Automated or staffStaffContent flags (Section 4); a domain resembling a known brand; the same message sent by many new Teams
3. PauseAll sending, or sending from one domain or key, is stopped. The API returns a clear error with the reason. The dashboard stays available.Automated or staffStaff onlyAn action level in Section 6; unresolved holds; a compromised key; a protective pause (Section 6)
4. SuspendThe account is restricted: no sending, no new keys or domains. Read access and export remain unless prohibited by law.Staff onlyStaff onlyA confirmed violation of this AUP; a pause that was not resolved; no answer to our questions; evasion
5. TerminateThe agreement ends under Section 6 of the Terms of Service.Staff onlyn/a (final), subject to appealSerious or repeated violations; fraud; a violation that threatens the platform

We may combine steps for one incident (for example, a hold on the affected Messages and a throttle on the rest of the Team), and we may move a Team down or back up the ladder as events develop.

9.3 Emergency path

Where there is evidence of phishing, malware, fraud, child sexual abuse material, credential compromise, an attack on the Service, or a risk that mail from Irisend will cause Amazon SES or a mailbox provider to restrict us, we may pause or suspend a Team immediately and without prior notice. We tell you afterwards, as set out below.

9.4 Notice and statement of reasons

Whenever we throttle, hold, pause, suspend, or terminate, we tell you through the dashboard and by email to the account owner, as soon as it is safe to do so (before the action where practicable, after it in an emergency). Where Art. 17 of the EU Digital Services Act applies, and in every other case where practicable, our notice is a statement of reasons describing: the facts and circumstances relied on; whether automated means were used to detect the issue or to decide on the action; the legal or contractual ground relied on (including the section of this AUP); and the redress options available. We may withhold details that would help a bad actor evade detection, or that would breach the privacy of a recipient or reporter.

9.5 Resuming and appeals

  • Resuming. To ask us to lift a pause or suspension, reply to the notice or email [email protected], and explain what happened, what you changed, and how you will prevent a repeat. A person reviews the request and decides. We may ask for proof of consent for the affected recipients, bounce and complaint breakdowns, or sending-domain ownership evidence (Section 2).
  • Appeals. If you believe an action was a mistake, or disagree with a suspension or termination, email [email protected] with the subject "Appeal" within 30 days of the notice. A person reviews it, where possible someone other than the person who made the original decision, and we aim to answer within 5 business days. We may keep the restriction in place during the review where lifting it would risk harm to recipients or to the platform. If we find an error, we restore your access and correct the record.
  • Other remedies. Appeals do not affect your right to contact your data protection authority, to use any out-of-court dispute settlement body that applies, or to bring a claim before the competent court under Section 14 of the Terms of Service.
  • Fixes we expect. The usual route back is: stop the offending sending, remove the unlawful recipients, fix list collection (for example by adopting double opt-in), authenticate your domains (Section 7), and show that consent evidence exists.

9.6 Reports to authorities

Severe violations (phishing, malware, child sexual abuse material, fraud) are reported to the relevant authorities where required by law, and we may preserve and share evidence for that purpose.

10. Reporting Abuse

To report suspected spam, phishing, or other abuse originating from the Service, email [email protected] (or [email protected]) with full headers and, if available, the offending Message, or use the report form at /report-abuse, which accepts the same information. Messages sent through Irisend carry identifiers we can use to trace them to the responsible Team; sending us the full headers (or the email ID) is the quickest route. A notice should include: (a) an explanation of why you consider the content illegal or abusive; (b) a clear indication of its location (e.g., the message ID, sending domain, and timestamp); (c) your name and email address; and (d) a statement of your good-faith belief that the information provided is accurate. Where the EU Digital Services Act (Regulation (EU) 2022/2065) Art. 16 applies, we will acknowledge receipt of your notice without undue delay and notify you of our decision. The abuse mailbox is monitored by a person; we aim to acknowledge abuse reports within 24 hours and act on confirmed violations promptly. We may forward a notice, with the reporter's contact details removed where the law allows, to the account holder so that they can respond.

11. DMCA / Copyright Complaints

To report content sent or hosted via the Service that infringes your copyright, send a notice to [email protected] including: (a) identification of the copyrighted work; (b) identification of the infringing material and its location; (c) your contact information; (d) a statement of good-faith belief that the use is unauthorized; (e) a statement, under penalty of perjury, that the notice is accurate and you are authorized to act; and (f) your physical or electronic signature. We will respond in accordance with the DMCA (17 U.S.C. § 512) and equivalent notice-and-takedown regimes elsewhere, and may forward your notice to the account holder.

12. Changes

We may update this AUP from time to time; material changes will be notified per Section 13 of the Terms of Service. Changes to thresholds and trust-tier criteria that tighten limits will take effect for all customers at the same time.

13. Contact

Abuse reports: [email protected]. General questions and appeals: [email protected].


This document is a draft template and does not constitute legal advice. See research/legal.md for review recommendations before publishing.