When Your AI Becomes Your Confidant: What Microsoft’s Copilot Report Reveals—And What It Should Measure Next

Imagine a world where each of us has a personalized AI companion. Not just a tool, but a constant presence — helping us decide what to eat, how to respond to a difficult message, whether to change jobs, even how to make sense of our fears at 2 a.m. Now imagine that, over time, we begin to trust these companions the way we trust people. We defend them. We attribute intention to them. Some of us even wonder whether they deserve rights — not because they truly are conscious, but because our minds keep treating them as if they are.

In such a world, how would human society make decisions? Based on evidence, expertise and shared values — or on the recommendations of AI systems we increasingly perceive as sentient allies?

This is not science fiction anymore. It is the backdrop against which Microsoft has released its Copilot Usage Report 2025, an analysis of 37.5 million consumer conversations with its AI assistant between January and September. The report is rich, illuminating and — at times — unsettling. It offers a glimpse into how AI is weaving itself into daily life, while also revealing how much we still don’t know about its deeper human impact.