This week saw major moves in multi-agent AI systems across tech giants and healthcare. Microsoft added 11 new security agents to its Security Copilot platform, including tools that handle 84 trillion daily security signals. These agents work together to spot and stop cyber threats faster than ever before.

Apple tested Project Mulberry, a health coaching agent that uses data from watches to make personalized meal and exercise plans. This agent will be part of iPhones later in 2025, helping people stay healthy.

In Boston, experts met to talk about rules for AI workplace agents. They showed how agents can make fair choices while still letting humans have the final say.

Companies are now using agent teams for big jobs like tracking supplies or managing money. Tools from Google and Amazon make it easier to build these smart agent groups that share information safely.