From Dreamforce: How can you effectively orchestrate and govern AI agents so you can safely trust...

Mila D'Antonio is an analyst at Omdia, covering Customer Engagement Intelligence. She just posted a Dreamforce preview on LinkedIn and her key takeaway was that Salesforce will spend less time convincing enterprises that agents are possible and more time explaining how millions of agents can successfully and safely operate together. She noted orchestration will be critical and asked a series of great questions. I’d love to hear what some of your answers might be:

How do agents discover and collaborate with other agents?

How is context shared?

How are permissions inherited? 

What happens when an agent behaves unexpectedly? 

Where does human oversight enter the process? 

And how does an enterprise maintain control as agents become increasingly autonomous?