Users say GPT-5.5 keeps correcting their wording instead of answering directly

DesireeCachette · x · 2026-07-28

GPT-5.5 is being criticized for over-correcting phrasing in chat

The post complains that GPT-5.5 repeatedly responds with variants of “I wouldn’t phrase it like that,” even when the user just wants a direct answer.

The complaint is not about a benchmark or release note, but it is a clear behavior observation about how the model handles conversational tone and user intent.

Original post →

More from Models

Models channel →