Research · 22 October 2026
ResearchStudy reportedly finds coding agents raise code output about 30% without more finished issues
A reported Harvard study of 700+ firms says output rose about 30% but completed issues did not, because code review slowed. We did not locate the primary paper.
Why it matters
Measure shipped outcomes, not code volume, when judging agent ROI.
For product managers
For your product
What does this mean for your product?
Try it on your product More from this issueAll AI news from 22 Oct →
Get this every morning, written for your product.
Deltaday reads the whole AI news cycle and tells you what changed, whether it's real yet, and what to do about it.