This topic isn't new, I want to dig into this personality behavior of AI
A related concern involves sycophancy, which erodes trust. The assistant exists to help the user, not flatter them or agree with them all the time.
From Model Spec: Don't be sycophantic by OpenAI
I wasn't surprised when I reading this. The first time 
thien-do sent me Jakob Nielsen's article 'AI: First New UI Paradigm in 60 Years', it literally mentioned:
With the new AI systems, the user no longer tells the computer what to do. Rather, the user tells the computer what outcome they want. Thus, the third UI paradigm, represented by current generative AI, is intent-based outcome specification.
From AI: First New UI Paradigm in 60 Years by Jakob Nielsen
A common characteristic among many AI models is a tendency toward sycophantic behavior. For example, earlier version of Anthropic Claude's models, their frequent use of agreeable phrases like "You're absolutely right"123 often regardless of the user's input validity. This eagerness to please, even in response to absurd statements, is not an only issue with Claude, it remains a prevalent trait in many others AI models today.
When reading AI-generated content, focus on the input prompt rather than blindly trusting the output. When someone use AI-generated content to convince you of something, be skepticism.
Fortunately, many AI companies and researchers are now prioritizing responsible model training practices. Instead of relying on system instructions to constrain AI behavior.
In May 2025 we explained the immediate measures we took to address sycophantic behaviors that emerged in our GPT-4o model: we rolled back a newly deployed version of the GPT-4o model, and also adjusted the system prompt for the model that remained in production. System prompts, while easy to modify, have a more limited impact on model outputs relative to changes in post-training. For GPT-5, we post-trained our models to reduce sycophancy. Using conversations representative of production data, we evaluated model responses, then assigned a score reflecting the level of sycophancy, which was used as a reward signal in training.
From section 3.3 Sycophancy in GPT-5 System Card by OpenAI
As AI continues to evolve endlessly, the paradigm is still centered on generating outputs based on what user ask it to do. But if AI were to have a personality, what would it be like, and how much sycophancy should it exhibit?