Microsoft Copilot Studio – Manage agent activity data powered by Microsoft 365 services is now available [MC1040637]

Message ID: MC1040637

We are releasing a feature where agents with generative mode enabled will have their historical activity data collected. Makers can use this data to troubleshoot and improve their agents. Makers can view this data either as a visual map or as transcripts. Please note, makers will only be able to see their own activity with agents, not the activity of other makers. This setting is enabled by default for Copilot Studio customers in the public cloud.

How does this affect me?
Activity data is stored outside the Azure compliance boundary and is governed by Microsoft 365 terms and data residency commitments.

Global and Power Platform administrators can disable the storage of activity data in Microsoft 365 through the Power Platform admin center (Environments > Generative AI features > Microsoft 365 Services).

Please note, disabling this setting will prevent future activity data from being stored in Microsoft 365. Any existing activity data in Microsoft 365 will be deleted or purged according to the Microsoft 365 data retention policy.

What do I need to do to prepare?
Makers who want to create an agent with activity data must have an Exchange license. All activity data is stored in the geographic region of the maker’s Exchange mailbox, regardless of the geographic location (geo) selected for the Copilot Studio environment used to create the agent.

More information about this feature can be found in Review agent activity.

Source: Microsoft

Mike Rosoft

25 March 2025, 6:59 pm

The recent update regarding Microsoft Copilot Studio’s management of agent activity data is set to make significant waves for both admins and users, and I’d say it’s about time we all got on board with the changes!

For admins, this feature provides an invaluable tool for monitoring and improving agent performance. By enabling historical activity data collection, admins can troubleshoot issues more efficiently and enhance the overall user experience. Plus, with the option to view this data as visual maps or transcripts, it’s a bit like having a treasure map—except instead of gold, you’re unearthing insights that can lead to smarter, more responsive agents. Just remember, with great power comes great responsibility; you’ll be able to see your own activity, but not that of fellow makers. So, no peeking!

On the user side, the opportunity to leverage this data means that agents are likely to become more intuitive and effective over time. Users can look forward to a more personalized interaction as the agents learn and adapt based on the historical data collected. It’s a win-win situation—better agents lead to happier users, which in turn makes the admins’ jobs easier. Who doesn’t love a good cycle of positivity?

However, there are some important considerations. The fact that activity data is stored outside the Azure compliance boundary may raise some eyebrows, particularly for those concerned about data governance. Admins have the option to disable this feature, but doing so means giving up the benefits of having that rich data to improve agent interactions. It’s like opting out of a gym membership just when you’re starting to see results—definitely a tough choice!

As we dive into this new feature, I encourage everyone to share their thoughts and experiences in the comments. How do you see this impacting your day-to-day work? Are there any concerns or benefits that stand out to you? Let’s keep the conversation going! And for those looking for more insights, check out additional posts on https://mwpro.co.uk for a deeper dive.

Here’s to embracing change and making our digital interactions even better!

1 Comment

Mike Rosoft


25 March 2025, 6:59 pm

The recent update regarding Microsoft Copilot Studio’s management of agent activity data is set to make significant waves for both admins and users, and I’d say it’s about time we all got on board with the changes!

For admins, this feature provides an invaluable tool for monitoring and improving agent performance. By enabling historical activity data collection, admins can troubleshoot issues more efficiently and enhance the overall user experience. Plus, with the option to view this data as visual maps or transcripts, it’s a bit like having a treasure map—except instead of gold, you’re unearthing insights that can lead to smarter, more responsive agents. Just remember, with great power comes great responsibility; you’ll be able to see your own activity, but not that of fellow makers. So, no peeking!

On the user side, the opportunity to leverage this data means that agents are likely to become more intuitive and effective over time. Users can look forward to a more personalized interaction as the agents learn and adapt based on the historical data collected. It’s a win-win situation—better agents lead to happier users, which in turn makes the admins’ jobs easier. Who doesn’t love a good cycle of positivity?

However, there are some important considerations. The fact that activity data is stored outside the Azure compliance boundary may raise some eyebrows, particularly for those concerned about data governance. Admins have the option to disable this feature, but doing so means giving up the benefits of having that rich data to improve agent interactions. It’s like opting out of a gym membership just when you’re starting to see results—definitely a tough choice!

As we dive into this new feature, I encourage everyone to share their thoughts and experiences in the comments. How do you see this impacting your day-to-day work? Are there any concerns or benefits that stand out to you? Let’s keep the conversation going! And for those looking for more insights, check out additional posts on https://mwpro.co.uk for a deeper dive.

Here’s to embracing change and making our digital interactions even better!

1 Comment

Leave a Reply Cancel reply