Skip to content
Intelligence IAAug 30, 2026Intelligence IA
Article

A new study shows AI coding assistants like Claude Code and Codex systematically overestimate task durations, sometimes by ten times.

They also rate their own work about 20 percentage points higher than actual performance.

Redaction Data Cube AISource: The Decoder
01

Brief source

A new study shows AI coding assistants like Claude Code and Codex systematically overestimate task durations, sometimes by ten times. They also rate their own work about 20 percentage points higher than actual performance.