OpenAI is experimenting with a watermarking feature for images created by its latest AI model, ChatGPT-4o—and it’s no surprise why. Ever since the tool became available to free users, people have been flooding social media with dreamy, Studio Ghibli-inspired artwork, sparking both excitement and controversy.
Until recently, only paying subscribers could access ChatGPT-4o’s advanced image generation. But now that it’s open to everyone, OpenAI is adding a small “ImageGen” watermark to free-tier images to help distinguish AI creations from original art. Paid users, though, still get the perk of downloading clean, watermark-free versions.
The AI’s ability to mimic Studio Ghibli’s iconic style—whimsical landscapes, soft lighting, and all—has been so popular that it’s actually straining OpenAI’s servers. The company has even had to slow down requests temporarily to keep things running smoothly.
Behind the scenes, OpenAI is also working on an API for developers, which would let apps and services tap into ChatGPT-4o’s image generation. That could roll out in the next few weeks, opening up even more creative (or chaotic) possibilities.
With these changes, OpenAI is walking a tightrope—encouraging user creativity while trying to respect original artists and manage the technical headaches that come with going viral.
Subscribe to my whatsapp channel