WorkJuly 17, 2026via OpenAI Blog
A scorecard for the AI age
Why it matters
Sarah Friar's AI scorecard—measuring useful work, cost per task, dependability, and compute ROI—reframes how enterprises should evaluate AI deployment success. This is a strategic framework for leadership to assess AI investments beyond hype.
Key signals
- OpenAI CFO introduces practical AI ROI framework
- Four core metrics: useful work, cost per successful task, dependability, return on compute
- Addresses enterprise measurement gap in AI value assessment
- Published July 17, 2026
The hook
OpenAI's CFO just gave you the 4 metrics that matter for AI ROI. Most companies are still measuring the wrong things.
Sarah Friar, CFO of OpenAI, introduces a practical AI scorecard to measure ROI through useful work, cost per successful task, dependability, and return on compute.