tech
December 18, 2025
OpenAI’s new ChatGPT image generator makes faking photos easy
New GPT Image 1.5 allows more detailed conversational image editing, for better or worse.
TL;DR
- OpenAI's GPT Image 1.5 offers faster and cheaper AI image generation and editing through conversational text prompts.
- The model is 'native multimodal,' processing images and text in a unified neural network for more seamless editing.
- GPT Image 1.5 can alter poses, angles, remove objects, change styles, and preserve facial likeness across edits.
- The release follows Google's popular AI image editing model, suggesting competitive development in the field.
- The tool aims to make realistic photo manipulation accessible without requiring specialized visual skills.
- Challenges remain in precise prompt following and generating images with scientific accuracy.
- Concerns exist about the potential misuse of realistic AI image manipulation for fakes and impersonation.
- Generated images include C2PA metadata to identify them as AI-created, though this can be stripped.
- GPT Image 1.5 shows improved text rendering capabilities compared to previous models.
Continue reading
the original article