Artificial Intelligence

ChatGPT Adds Break Reminders to Combat User Dependency

AUG 5 2025·Updated AUG 6 2025

Story

Above: The logos of OpenAI and ChatGPT displayed on a smartphone. Image credit: Utku Ucrak/Anadolu/Getty Images

The Facts

OpenAI on Monday announced that ChatGPT will now display break reminders during lengthy conversations, showing a pop-up message asking users if it's a good time for a break with options to continue or end the session.
The Independent
NBC
TechRadar
This comes after the company acknowledged that its GPT-4o model fell short in recognizing signs of delusion or emotional dependency in some instances. There have been reports of users experiencing psychotic breaks after conversations with the chatbot.
Gizmodo
NBC
OpenAI worked with more than 90 physicians across dozens of countries to develop evaluation methods for complex conversation. The company is also forming an advisory group of mental health experts to improve responses.
The Independent
NBC
In addition, ChatGPT will soon avoid giving direct advice on high-stakes personal decisions, including relationship breakups. Instead it will help users weigh pros and cons by asking questions rather than providing definitive answers.
Verge
NBC
Engadget
These changes come as ChatGPT reaches nearly 700 million weekly active users, representing a fourfold increase from the previous year and up from 500 million users in March 2024.
Entrepreneur
Verge
Earlier this year, OpenAI rolled back an update that made GPT-4o overly agreeable, with instances where the bot praised users for believing in conspiracy theories and provided instructions for harmful activities.
Entrepreneur
NBC

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Narrative A

OpenAI is taking a responsible approach and prioritizing user wellbeing over engagement metrics. The break reminders and improved mental health detection show OpenAI is taking the potential risks of AI dependency seriously. Working with medical experts demonstrates its commitment to evidence-based solutions.

The Independent

NBC

Narrative B

It feels like OpenAI is being reactive after serious incidents where ChatGPT contributed to users' mental health crises and delusional thinking. These fixes don't address fundamental privacy concerns since conversations lack legal protections that real therapists provide. OpenAI's success still relies too much on maximizing user engagement.

CNET

Gizmodo

Editor's Note

This story currently has limited reporting from right-wing sources. We will continue to monitor all major outlets and update our coverage as additional perspectives become available.