---
title: "What Score Makes an AI Agent Pass? The \"Measurement Science\" Advanced by Microsoft and CMU"
date: 2026-10-24
source: http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6ade1d9e65d24c6c9d0dc7c25d3c7068&url=https%3a%2f%2fnote.com%2fturuto%2fn%2fne71184cfcccb%3fhl%3den&c=5975806949338905068&mkt=en-us
description: "AI is no longer just returning text; it is now using tools, executing multi-step processes, and acting on behalf of humans.However, what should we measure to determine if it is safe to \"use\" these AI ..."
---

# What Score Makes an AI Agent Pass? The "Measurement Science" Advanced by Microsoft and CMU

AI is no longer just returning text; it is now using tools, executing multi-step processes, and acting on behalf of humans.However, what should we measure to determine if it is safe to "use" these AI ...

*Published: 2026-10-24*