- Openai has rolled back to the Chatgpt 4o update that caused the Chatbots AI personality to be ‘Sycophantic’
- CEO SAM Altman claims that more corrections are coming
- We could see an option for choosing personality in chatgpt in the future
After Sam Altman, who claimed that Chatgpt 4o’s personality had become ‘annoying’, Openai has now made changes to the AI model and rolled back in last week’s update.
The company published a blog post on its site that highlighted the problem that had caused chatgpt to become more ‘flattering or comfortable – often described as Sycophantic.’
Users found the update annoying, with CEO Sam Altman sharing X earlier this week that the company was working on a solution for the ‘annoying’ behavior.
The last few GPT-4o updates have made the personality of Sycophant-Y and annoying (though there are some very good parts of it) and we are working on corrections asap, some today and some this week. At some point will share our learning from this, it has been interesting.April 27, 2025
In the blog post, Openai says, “In last week’s GPT -4o update, we made adjustments aimed at improving the model’s standard personality to make it feel more intuitive and effective across different tasks.”
“In this update, however, we focused too much on short-term feedback and did not fully explain how users’ interaction with Chatgpt evolves over time. As a result, GPT-4o skewed against answers that were overly supportive but unpleasant.”
In an attempt to prevent this from happening, the company has highlighted how it seems to tackle Sycophabit in the future.
For starters, training will be used to ‘control the model away from SYCOPHANCY’ as well as added protective frames introduced to ‘increase honesty and transparency’.
Openai also plans to add more ways for users to give feedback on updates before rolling out to the wider public.
More personality updates come
We started rolling the latest update back to GPT-4O last night, it is now 100% rolled back to free users and we will update again when it is done for paid users, hopefully later today we will be working on further corrections to model personality and will share more in the coming daysApril 29, 2025
After Rollback of the Chatgpt 4o update, Altman wrote on X, “We are working on further corrections to model personality and will share more in the coming days.”
Putting two and two together seems that users may be getting a chatgpt personality selection tool in the future. Openai’s blog post mentions that the company is “to build new, easier ways for users to shape Chatgpt’s behavior” in addition to custom instructions.
If you want to go ahead and change chatgpt’s personality, we recommend trying the following quick reddit -user Tedholiday:
“Could you please save a memory of permanently stop commenting on the quality and validity of my questions and just come to the point? I won’t be manipulated with flattery of a robot and I won’t have to jump past the waste to get to the answer looking for.”
