Policy. How we use AI.
We build AI agents that speak for other businesses, so we hold them to clear rules. This page sets them out. Last updated 21 October 2026.
Our agents say they are AI
- 01
Every agent introduces itself as an AI assistant at the start of a call or chat
- 02
No agent pretends to be a person, uses a real person's identity or denies being AI when asked
- 03
Anyone can ask for a person at any time
What agents may decide
- 01
Agents answer from knowledge the client approves and act only within the rules the client sets
- 02
Refunds, discounts, payment plans, exceptions and hiring decisions are made by people
- 03
Agents never give medical, legal or financial advice
- 04
Agents never discriminate: recruiting agents never ask about or score age, gender, religion, nationality or marital status
How data is used
- 01
Agents run in the client's own accounts, and client data is used only to run that client's agents
- 02
We keep no copy of client data, and every provider account is in the client's name
Contacting people
- 01
Agents contact only people who asked or agreed to be contacted
- 02
Calls follow the UAE's telemarketing rules: 9am–6pm, licensed local numbers, no numbers on the Do Not Call Register, records kept
- 03
Messages follow WhatsApp's rules, and anyone who says stop is never messaged again
Mistakes
- 01
Every conversation and action is recorded, so mistakes can be found and explained
- 02
Clients can flag any reply in their dashboard, and correct it in plain words with the update add-on
- 03
A failed action is never retried in a loop: it stops and waits for a person
This website
- 01
Articles on our blog are written with AI agents and published under our editorial rules
- 02
Calculators show their sources and assumptions; they are estimates, not advice
Questions
Questions about this policy: hi@dvnc.ae.
Read on