Legal
AI & Automation Policy
How the AI in Kijani works, what it can do on its own, and how you stay in control.
- Version
- 1.0
- Last updated
- 6 October 2026
- Effective
- [EFFECTIVE DATE]
Draft for legal review. This document has not yet been reviewed by a lawyer. Highlighted [ITEMS] are decisions still to be made.
1.What the AI creates
Kijani’s AI may create: emails and WhatsApp messages, replies to customers, social posts and captions, video scripts and videos, research summaries about businesses, and lead information such as a business’s contact details found on its own pages.
AI output may contain mistakes. Kijani checks messages against what it actually found and holds back claims it can’t support, but you should still review important content before it goes out.
2.No guaranteed results
The AI does not guarantee leads, sales, replies, revenue or accuracy. Results depend on your offer, your market, the people contacted, and the services you connect.
3.Actions the AI can take on its own
Depending on how you configure it, the AI can act automatically: search for businesses on a schedule, run the goals and plans you start, send emails and WhatsApp messages, reply to conversations, send follow-ups, make content and post it to your connected pages.
You control this in Settings → AI controls:
- Permissions: turn each kind of automatic work on or off.
- Approval mode: Manual approval (the AI prepares, you approve), Smart approval (messages to strong-match leads go out automatically, the rest wait for you), or Autonomous (permitted actions happen automatically). Autonomous mode is never on by default and needs your explicit authorisation, which is recorded.
- Limits: a daily maximum of emails and WhatsApp messages, and the hours when messages may be sent.
- Pause AI: one switch that stops all automatic work at once until you resume it.
- Activity: a log of what the AI did and what is waiting for you.
These controls are enforced by our servers, not only shown in the app.
4.Your responsibility
You are responsible for the instructions you give, how agents are configured, the accounts you connect, who is contacted, campaign settings, the content that goes out, and the permissions you grant.
5.Human oversight
We recommend approving messages before they are sent, especially at the start of a plan, for new audiences, and for anything sensitive. New plans ask you to approve their first messages before anything is sent. Replies the AI can’t back up from your knowledge base are drafted and held for you, and a conversation can be handed to a person at any time.