Mailday AI Use Policy
Last updated: 2026-05-13
Mailday's AI Features are designed to help members and workspaces move faster — drafting emails, preparing invoices, scheduling content, replying to comments, and recommending payouts. This Policy explains what the AI does, what it doesn't do, and your responsibilities.
This Policy supplements, and does not replace, our Terms of Service, Privacy Policy, and Acceptable Use Policy.
1. Our AI provider
We use AI models from Anthropic, PBC (anthropic.com) to power AI Features. We may add or substitute providers over time; the current list is in our Subprocessor List.
2. What AI can do at launch
Depending on your plan and configuration, AI may:
| Capability | Default behavior |
|---|---|
| Drafting emails | Draft only — you choose whether to send |
| Preparing invoices | Draft only — you choose whether to send |
| Recommending payouts | Where you enable auto-payouts, the system fires payouts up to your configured AI payout cap; amounts above the cap route to your Workspace Admin for approval |
| Social post drafting | Draft, schedule, and (where you have enabled it per platform) auto-publish |
| Replying to social comments | Draft only — you choose whether to post |
| Categorizing inbox and assigning workflow status | Automatic; you may archive statuses |
3. Human review and approvals
3.1 Sensitive actions
The following actions require Workspace Admin approval before execution:
- Payouts above a configured threshold
- Invoices above a configured threshold
- Bulk deletions of workflows, messages, or media
- Public posting of content flagged for review
3.2 Approval thresholds
Workspace Admins can set thresholds in Workspace settings. We provide reasonable defaults; you should configure thresholds appropriate to your business.
3.3 Workflow status overrides
You may archive AI-generated workflow statuses. Other manual overrides of AI-generated statuses are not supported at launch.
3.4 Disabling AI
Workspace Admins (or individual Members) can disable AI Features at the Workspace level, or pause AI temporarily, from settings. Disabling AI may reduce or eliminate certain functionality.
4. Your responsibilities
You are responsible for everything that is sent, posted, paid, or invoiced from your Account, including AI-generated and AI-initiated actions. This means:
- Review before you authorize. For drafts that require your send/publish action, review the content first.
- Set sensible thresholds. Configure approval thresholds for payouts and invoices to fit your risk tolerance.
- Disclose AI use where required. Some platforms and jurisdictions require you to disclose AI-generated content. Comply with those requirements.
- Comply with the FTC Endorsement Guides and equivalent rules in your jurisdiction, including disclosure of company relationships.
- Don't use AI to deceive. Don't impersonate real people, generate sexual content involving minors, harass, or otherwise violate our Acceptable Use Policy.
5. Data and training
- Customer data is not used to train AI models by Mailday or, per our contractual arrangement, by Anthropic.
- Prompts, AI outputs, and AI action logs are retained per the schedule in our Privacy Policy.
- Where you disable AI Features, we cease sending Workspace data to AI providers for those features.
6. Limitations
AI outputs can be wrong, outdated, or inappropriate. AI may:
- Misinterpret instructions
- Invent details ("hallucinate")
- Misclassify messages or workflows
- Generate content that is off-brand or off-tone
- Misidentify amounts, parties, or dates in invoices and payouts
You are responsible for verifying AI outputs and actions. Mailday is not liable for AI-generated content or actions you authorized, including auto-publish actions you have enabled. See Section 17 (Limitation of Liability) of the Terms of Service.
7. Viewer AI chat
Where you enable it, Viewers may chat with AI inside a Shared Report you have provided. AI responses are scoped to the data in that report. By enabling viewer chat, you authorize the AI to respond to questions using report data and acknowledge that viewer interactions are logged for security and quality. You may review what each Viewer asked from the report's viewer view in your dashboard.
8. Reporting AI issues
If an AI output or action causes harm or appears inappropriate, contact support@mailday.ai. For urgent issues (e.g., an erroneous public post or an erroneous payout), also use the in-product Pause AI control to temporarily disable all AI features in your workspace.
9. Changes
We may update this Policy as AI capabilities evolve. Material changes will be announced in-product or by email.
Contact
- support@mailday.ai
- privacy@mailday.ai
Last updated: 2026-05-13