AI Briefing
KO

ChatGPT Image Model Inserts Text

·2026.04.26 01:44

Key point

ChatGPT Images 2.0 added text to an image that wasn't in the prompt.

Details

Scott jla reacted to a generated image, saying "I need to run more tests."

The prompt was to depict a horse riding an astronaut, the astronaut riding a pelican, and the pelican riding a bicycle. However, ChatGPT Images 2.0 added a "WHY ARE YOU LIKE THIS" sign into the image that wasn't part of the request, and Simon Willison checked the shared conversation and result to verify that the model had added the text on its own.

multimodalchatgptimage-generationtext-rendering
Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Related

RedditSenseNova-U1.5-8B-MoT Released: Unified Text-to-Image and Understanding Model Initialized from Qwen3RedditInternLM Releases Intern-Decision 4B and 0.8B Multimodal Structured Decision ModelsKarrotFake Customer Service Pages Induce ChatGPT and Gemini to Display Scam NumbersElevenLabsElevenLabs Launches Image, Video, and Templates APIs for Automated Creative ProductionTLDR AIOpenAI Prepares New $500/Month ChatGPT Pro Max Plan with 'Fastest' InferenceHacker NewsClass Action Suit Alleges OpenAI Routed ChatGPT Conversations to Human Reviewers Without Clear Disclosure