Back to News
AI
AI Coding Agents Write 30% More Code – and Resolve Zero Additional Issues, Harvard Study Finds
Forkast.news·October 11, 2026·1 min read
AI Summary
A Harvard study reveals that AI coding agents generate 30% more code than human developers but fail to resolve any additional issues, highlighting a disconnect between code volume and problem-solving effectiveness. The findings suggest that enterprise leaders focusing on code output as a success metric may be overlooking the actual value and quality of AI-assisted development work.
From the source
Enterprise technology leaders are currently caught in a measurement trap. When evaluating the impact of AI coding agents, the industry has largely focused on the volume of code produced. It is an easy metric to track, but it is increasingly clear that it is t…
The full text couldn't be loaded here (the source may require a subscription).
View original at Forkast.newsKeep reading
All AI is ‘compromised’ and has fewer rules than milk, Microsoft bosses say - The Independentindependent.co.uk · 2h agoBritain steps up national defences against growing AI risks - GOV.UKgov.uk · 2h agoBouygues Telecom at Dataiku Summit Paris 2026: Agentic AI for Network Excellencecorporate.bouyguestelecom.fr · 2h agoYour next bank manager could be AI: How Bradesco and Santander are preparing and whether banking apps will disappearseudinheiro.com · 2h ago
Was this useful?