Since they're generated for every interaction regardless, it might be worth considering what Program you enable AI insights on. For instance, if you have a queue where there's possibly no benefit to Insights, I wouldnt assign an AI Insight enabled program to that queue. Any queues that might take on more complex enquiries or queues where performance/improvement is the focus, look to utilise it on those.
Although it brings the question, should AI Insights be generated for every interaction or should we be more resource conscious and only have it work for those interactions that are reviewed? Surely if a contact centre that might generate for example 30,000 calls a month, but only 5% of those calls are reviewed. There's 95% of additional AI resources spend on interactions that will never get any benefit from it. If the system is generating for every interaction, wouldn't it potentially generate just as fast as opening the interaction if the system wasn't working for every interaction? Just a thought, would be curious to understand why it's built this way.
Original Message:
Sent: 03-24-2026 16:18
From: Phaneendra Avatapalli
Subject: Does Supervisor Copilot AI Insights Use Tokens If Agent Copilot Is Already Enabled?
Hi Caleb,
From a token perspective, consumption is based on when the summary or insight is generated, not when it's viewed.
So if summaries are generated automatically for interactions, they will count towards token usage regardless of whether someone opens them later.
Hope this helps happy to be corrected if I've misunderstood anything
------------------------------
Phaneendra
Technical Solutions Consultant
Original Message:
Sent: 03-24-2026 10:44
From: Caleb Koehler
Subject: Does Supervisor Copilot AI Insights Use Tokens If Agent Copilot Is Already Enabled?
I wanted to follow up with a clarification on how token consumption works, specifically regarding this line:
"AI Summary and Insights: 50 AI summaries/insights per token."
Do summaries generate automatically for every interaction, or does a summary only count against our usage when someone actually opens and views it?
I want to make sure that, from a billing perspective, we're not being charged for summaries on every incoming interaction-only for the ones we intentionally choose to view.
Thank you!
------------------------------
Caleb Koehler
SYSTEM AND DEVELOPMENT ANALYST 1
Original Message:
Sent: 03-17-2026 00:10
From: Phaneendra Avatapalli
Subject: Does Supervisor Copilot AI Insights Use Tokens If Agent Copilot Is Already Enabled?
Thanks for the detailed explanation, that helps clarify things.
So if I understand correctly:
• If Agent Copilot is enabled and generating summaries, then Supervisor Copilot can reuse those summaries and insights without consuming additional AI tokens.
• But if Supervisor Copilot generates summaries independently, then the normal token consumption model (50 summaries/insights per token) applies.
In our case, we already have Agent Copilot enabled, so the plan would be to ensure Agent Copilot summaries are active and then allow supervisors to leverage those summaries through Supervisor Copilot.
That seems like a good way to get value from the insights without increasing token usage.
Appreciate you sharing the real implementation examples as well.
------------------------------
Phaneendra
Technical Solutions Consultant
Original Message:
Sent: 03-16-2026 22:58
From: Phaneendra Avatapalli
Subject: Does Supervisor Copilot AI Insights Use Tokens If Agent Copilot Is Already Enabled?
Hi everyone,
I have a question regarding Supervisor Copilot AI summaries and insights in Genesys Cloud.
Our organisation already has Agent Copilot enabled, and my understanding is that the AI processing used for Agent Copilot is already covered as part of that capability.
I've been advised that if Agent Copilot is enabled, then Supervisor Copilot summaries and insights may not consume additional AI tokens, as they can rely on the same underlying AI processing.
Could anyone confirm if this is the correct behaviour?
Specifically:
• If an organisation already has Agent Copilot enabled, do Supervisor Copilot AI summaries and insights still consume tokens?
• Or are they effectively covered by the Agent Copilot capability?
Just trying to understand the token consumption model before enabling this more broadly for supervisors.
Thanks in advance!
#AICopilot(Agent,SupervisorAdmin)
------------------------------
Phaneendra
Technical Solutions Consultant
------------------------------