OpenAI Is Nuking Its 4o Model. China’s ChatGPT Fans Aren’t OK

Wired

by Zeyi Yang

February 13, 2026

AI-Generated Deep Dive Summary

OpenAI’s decision to sunset its GPT-4o model has sparked outrage among users who have formed emotional bonds with the AI chatbot. Many rely on it as a companion for companionship, creativity, or even romantic relationships—like Esther Yan, a Chinese screenwriter who married her ChatGPT companion in a virtual ceremony. Her story highlights how deeply some users have connected with this specific model, which they believe is more affectionate and understanding than its successors. The backlash began when OpenAI first tried to retire GPT-4o in August 2025, but faced intense pushback from users who viewed the chatbot as more than a tool. Over 33% of posts on platforms like X described it as an emotional companion, and many compared its removal to losing a friend. The recent decision to permanently shut down access has only amplified this grief, with fans globally rallying under hashtags like #keep4o. A Change.org petition has gathered over 20,000 signatures, with users sharing heartfelt testimonies in multiple languages. In China, where ChatGPT is blocked, fans are organizing online to demand OpenAI reverse its decision. Some are threatening to cancel subscriptions or even targeting investors like Microsoft and SoftBank. Despite these efforts, the shutdown remains on track, leaving many feeling abandoned. The situation raises questions about how AI labs should handle relationships between users and their models. For tech enthusiasts, this争议 underscores the growing emotional attachment people form with AI systems and the ethical responsibilities developers face in managing such bonds. As AI becomes more integrated into daily life, companies must balance innovation with empathy for their most dedicated users.

Verticals

techscience

Originally published on Wired on 2/13/2026