AI Assistant Automatically Reactivating After Human Intervention
Status: Closed · Asked by Gaurav Arora on · 0 views
Hello Arshil,
Thank you for the clarification.
The issue is not limited to a single chat. We are experiencing the same behaviour across essentially every conversation where a human agent intervenes.
The behaviour is:
- AI Assistant is active.
- Customer sends a message.
- Human agent intervenes and manually responds.
- AI Assistant appears to be inactive/intervened.
- Customer sends the next message.
- AI Assistant automatically starts responding again.
- This happens repeatedly across different chats.
Our expectation is that once a human agent intervenes, the AI Assistant should remain disabled for that particular conversation until the human explicitly re-enables it.
Since this is happening consistently across chats, could you please check whether there is any account-level setting, Flow configuration, automation, trigger, or current Chatflow behaviour that is causing the AI Assistant to be reassigned/reactivated whenever a new customer message arrives?
We can certainly provide a few chat IDs for investigation, but since the behaviour is reproducible across multiple conversations, it appears to be a systematic issue rather than an isolated chat-status change.
Please specifically confirm:
- What event causes the AI Assistant to become active again after intervention?
- Is there any setting that permanently disables AI after human intervention?
- Is the Intervened status supposed to survive subsequent customer messages?
- Can we configure the AI Assistant so that human intervention = AI permanently paused until manually resumed?
- If this is currently not possible, is there a workaround using Pabbly Connect, Flow Builder, webhook, API, or another mechanism?
This behaviour is currently disrupting our human sales conversations because the AI can take over again immediately after our sales team has intentionally intervened.
Please treat this as a reproducible product behaviour/issue rather than an issue with one particular chat.
Regards,
Gaurav Arora
Hello Gaurav,
Thank you for sharing the details and the WhatsApp number where you are experiencing this issue.
However, when we checked the provided WhatsApp number, we could not find any AI Assistant associated with it.
Could you please share the correct WhatsApp number so that we can test and identify the issue and assist you further?
Hello Arshil,
Thank you for the clarification.
The issue is not limited to a single chat. We are experiencing the same behaviour across essentially every conversation where a human agent intervenes.
The behaviour is:
- AI Assistant is active.
- Customer sends a message.
- Human agent intervenes and manually responds.
- AI Assistant appears to be inactive/intervened.
- Customer sends the next message.
- AI Assistant automatically starts responding again.
- This happens repeatedly across different chats.
Our expectation is that once a human agent intervenes, the AI Assistant should remain disabled for that particular conversation until the human explicitly re-enables it.
Since this is happening consistently across chats, could you please check whether there is any account-level setting, Flow configuration, automation, trigger, or current Chatflow behaviour that is causing the AI Assistant to be reassigned/reactivated whenever a new customer message arrives?
We can certainly provide a few chat IDs for investigation, but since the behaviour is reproducible across multiple conversations, it appears to be a systematic issue rather than an isolated chat-status change.
Please specifically confirm:
- What event causes the AI Assistant to become active again after intervention?
- Is there any setting that permanently disables AI after human intervention?
- Is the Intervened status supposed to survive subsequent customer messages?
- Can we configure the AI Assistant so that human intervention = AI permanently paused until manually resumed?
- If this is currently not possible, is there a workaround using Pabbly Connect, Flow Builder, webhook, API, or another mechanism?
This behaviour is currently disrupting our human sales conversations because the AI can take over again immediately after our sales team has intentionally intervened.
Please treat this as a reproducible product behaviour/issue rather than an issue with one particular chat.
Regards,
Gaurav Arora
Pabbly Support:
Hello Gaurav,Thank you for sharing the details and the WhatsApp number where you are experiencing this issue.However, when we checked the provided WhatsApp number, we could not find any AI Assistant associated with it.Could you please share the correct WhatsApp number so that we can test and identif
Whatsapp Number is 8091355155
Hello Gaurav,
Thank you for reaching out to us and for sharing the details.
We checked your settings and found that under Settings > Inbox Settings, the AI Auto Reply option is enabled for All Contacts.
Because of this, whenever any contact sends a message, the selected AI Assistant is automatically assigned to that conversation, even if a human agent has previously intervened.
If you do not want the AI Assistant to automatically respond to contacts after human intervention, you can disable the AI Auto Reply option for All Contacts from the Inbox Settings.
Please try this and let us know if you need any further assistance.