Power Platform Community Forum Thread Details

Hello everyone,
I’m experiencing an issue in Copilot Studio Agent where AI-generated responses are being appended after the original answer in my topic. This behavior occurs only in the parent bot, while child bots do not show this problem.
Details:

The parent bot adds extra AI-generated text after the intended response.
Child bots display the topic answer correctly without any additional content.
The extra text seems to be generated automatically and is not part of my configured topic response.
I have already tried adding “End conversation” after the topic, but the issue persists.

Questions:

Why does this happen only in the parent bot?
Is there a setting or configuration to disable these additional AI-generated responses?
Could this be related to the parent bot’s AI settings or fallback behavior?

Any insights or solutions would be greatly appreciated!

Categories:

Topic creation & management

General topics

Hello KT-08120349-0,

Thanks for raising it in the Q&A forum.

This behavior is almost always tied to generative answers and orchestration settings that are enabled on the parent agent but not on the child agents.

In Copilot Studio, a parent agent can use generative answers / orchestration to “improve” or complete replies whenever the authored topic output is short, ends abruptly, or does not fully satisfy the detected intent.

Child bots often run in a more constrained mode (no or limited generative answers, no extra knowledge sources), so they just show your authored topic message and stop, while the parent continues with an additional AI-generated explanation or suggestion.

On the parent agent, the global Generative AI and Orchestration settings determine whether the bot is allowed to add extra AI text after your topic ends.
System/fallback topics (like generative “Fallback” or “Boost” behaviors) can also kick in after your topic’s last node, which is why adding an End conversation action alone may not fully stop the extra text.

What to change in the parent agent

In the parent agent, go to Settings → Generative AI (or AI capabilities) and either switch the response mode to a more restrictive option or turn off generative answers/orchestration for normal chit‑chat and unknown intents.
Under System topics (or AI/fallback settings), disable or limit generative Fallback / Boost / “answer on its own” behavior so the parent does not append extra AI content after a topic has already replied.
On the specific topic, open Topic → Settings / AI capabilities and turn off any “Create generative answers” or similar per‑topic AI options, then make sure the last node is a clear End current topic / End conversation node and retest in the parent bot.

If it helpful accept the answer.

Best regards,

Jerald Felix