OpenAI updates GPT-5 to be more ‘friendly’ again after users complain of cold responses

OpenAI has modified the personality of its newest model, GPT-5, to make it "warmer and friendlier" just a week after launch. This comes after users complained online that the AI model was too formal and was making robotic responses compared to earlier ones.

Sam Altman announces GPT-5 update

CEO Sam Altman said on X that user attachment to specific AI personalities had been stronger than the company expected.

“Most users should like GPT-5 better soon,” Altman said. “The real solution here remains letting users customise ChatGPT's style much more. We are working on that," he added.

After GPT-5 was launched on August 7, users flooded Reddit with complaints about the model's cold, corporate tone.

Users shared certain emotional behaviour from the bot, such as "Great question" or "great start", expressing disappointment with the new model.

Initially, OpenAI removed access to legacy models without warning, forcing users to check the new model, drawing huge criticism.

Most users should like GPT-5 better soon; the change is rolling out over the next day.

The real solution here remains letting users customize ChatGPT's style much more. We are working that! https://t.co/RfiYJ8AkEO

— Sam Altman (@sama) August 15, 2025

Also read: Bhumi Pednekar launches premium water brand BACKBAY priced at ₹200, internet mocks: ‘Bless her Bollywood bubble’ (startuppedia.in)

GPT-5 offers 3 new modes

The updated GPT-5 now offers users three distinct modes: "Auto," "Fast," and "Thinking," providing users with greater control over the AI's behaviour and response style.

"One lesson we've learned from the past few days is that we need to create a system that allows for more individual customisation of model personality," Altman said in response to the backlash online.

He admitted being surprised by users' strong attachment to specific AI models, calling it "different and stronger than the kinds of attachment people have had to previous kinds of technology".

Users gave innumerable feedback to Altman and members of the GPT-5 team during a Reddit ‘Ask me anything’ (AMA) session.

Many users expressed dismay with the new model and asked for GPT-4o to be brought back, especially those who had built their workflows around it.

Sam Altman’s OpenAI quickly reversed course, altering access to GPT-4o and raising rate limits to 3,000 messages per week for GPT-5's reasoning functionality.

Also read: Logistics costs to drop to single digits by 2026: Nitin Gadkari (startuppedia.in)