Trust / AI disclosure
Automation should be
visible and accountable.
01
Clear disclosure
Where an ORENFOLD system uses an AI assistant, the experience should clearly communicate that fact before collecting information or continuing a substantive conversation.
02
Approved boundaries
AI agents are configured around approved business knowledge, availability, service areas, escalation rules and prohibited claims. They should not invent technical advice, pricing or commitments.
03
Human involvement
A customer can request a human where the connected business supports it. Defined safety, uncertainty, value and policy signals may also trigger escalation automatically.
04
Data and consent
Collection should be minimised to what the workflow needs. Recording, messaging and outreach require appropriate disclosure, consent and legal review for the business, jurisdiction and provider involved.