“Reduce overly flattering language”: OpenAI changes GPT-6 for free ChatGPT users

“Reduce overly flattering language”: OpenAI changes GPT-6 for free ChatGPT users

OpenAI is changing GPT-6 in ChatGPT to give more concise answers and use less overly flattering language. The update applies to Free and Go users and people using ChatGPT without signing in.

For Plus and Pro subscribers, there is a narrower change: OpenAI says the update affects only the model ChatGPT can switch to after someone reaches GPT-6 message limits. The main GPT-6 model on those paid plans is unchanged.

The October 29 release note says the aim is to “reduce overly flattering language”. It does not provide example conversations or comparative test results, so it does not establish how often an individual will notice a difference.

Which ChatGPT users will get different responses?

The update concerns GPT-6 in the Chat experience. Free and Go account holders are included, as are people who use ChatGPT while signed out. OpenAI’s note does not describe a setting users need to enable.

Plus and Pro subscribers may encounter the revised behaviour if ChatGPT switches to its fallback after they hit the GPT-6 message limit. OpenAI says the main model for those plans is unchanged, so the note does not mean every conversation on a paid plan will sound different.

How this relates to the earlier GPT-6 rollout

OpenAI introduced GPT-6 more broadly in ChatGPT earlier in October alongside Intelligent UI, which can include visual or interactive elements in an answer. The new announcement concerns the affected model’s response style rather than a change to that interface.

The company says it is aiming for shorter and less flattering responses; that is not a promise that every answer will be brief or that factual mistakes will disappear. Users can still ask for more detail and check important claims.

Leave a Comment

Your email address will not be published. Required fields are marked *