Human intervention and contesting automated replies
How end-customers can request human review when a business uses Handlet auto-send.
Human intervention and contesting automated replies
Under UK GDPR Article 22, individuals may have rights when a decision about them is made solely by automated means and produces legal or similarly significant effects.
Handlet keeps conditional auto-send disabled by default. When a tradesperson enables it on Pro, they remain responsible for lawful basis, transparency to their customers, and providing a route to human review.
Stage 2 and meaningful human review
Before per-intent auto-send can unlock, operators build validated training evidence through draft review — approvals, edits, and correctness feedback per intent. This training path keeps operators in the loop before automation expands for a given message type. It does not remove the customer’s right to request human review after a message is sent.
For tradespersons using Handlet
Before enabling auto-send:
- Confirm your lawful basis for automated replies (contract necessity or explicit consent).
- Tell your customers that an AI-assisted reply may be sent automatically when you enable per-intent automation.
- Provide a clear contact route for human review (email or phone is sufficient).
Suggested notice to your customers
Some replies may be sent automatically using AI assistance when we have enabled that for routine messages. You can ask for a human review at any time by contacting us at [your support email or phone].
How customers request human review
Customers should contact the tradesperson or business directly using the contact details shown on quotes, invoices, websites, or message footers. Handlet does not decide complaint outcomes, refunds, or legal positions on behalf of the business.
If you are a customer and cannot reach the business, contact Handlet support at privacy@handlet.ai with the business name and approximate date of the message. We will route the request to the workspace owner where appropriate.
What Handlet never auto-sends
Even when auto-send is enabled, platform rules block:
- Risk, noise, unknown, or unsupported intents
- High-severity complaints (complaint auto-send requires separate opt-in and still blocks high severity)
- Reviews at two stars or below
- Quote, booking, or complaint sends when guardrails fail
Operators choose which eligible intents to automate with per-intent toggles. Disabled intents always require human approval.