Human in the loop, or human approval, means having a person approve what an artificial intelligence is about to do: send an email, launch a campaign, send an SMS. The AI prepares, the human decides.
How human in the loop works
- The AI assistant or agent prepares the action and puts it on hold, without executing anything.
- A person is notified, sees exactly what will go out (recipients, subject, message) and chooses: approve or refuse.
- If approved, the action goes out; if refused or not answered in time, nothing goes out.
- It can be tuned: approve every send, approve only campaigns, or switch to automatic with limits.
Why it matters for email marketing
- A sent email cannot be taken back: a wrong list, a wrong price or an awkward tone reaches all your contacts.
- Approval lets you start automating with an AI with confidence, then move to automatic sending once the results are good. It is not a brake: it is what lets you automate with peace of mind.
Human in the loop at Mailpro
- When an AI assistant goes through the Mailpro MCP server, every send waits for your approval by default: you approve in Mailpro or from the link in a notification email, without even signing in.
- Assistant by assistant, you can switch email or SMS to direct sending, with a daily cap, a log of what the assistant did and suspension of all sends in one click.
- The assistant connects with OAuth: you see its rights and can withdraw its access at any time.
Mailpro and AI assistants: approved or automatic sends, your choice. Start for free