ChatGPT Unlocks Image Creation for Free Users: A New Era of AI-Generated Art

The integrated image creation functionality of ChatGPT is now accessible to all users. OpenAI’s CEO Sam Altman mentioned last week that the company is postponing the introduction for free-tier users «for a bit longer,» as the demand for the tool exceeded their forecasts. However, the feature was made available to free users over the weekend, enabling them to create images directly within ChatGPT without needing to switch to OpenAI’s DALL-E. Before this, the tool was exclusively for Plus, Pro, and Team subscribers.

Altman had earlier indicated that free-tier users would be restricted to three images daily. Based on our trials, some accounts have been limited to just one image each day, while those able to generate multiple images may experience waits from several minutes to a few hours. One of the most favored applications of the tool we’ve observed so far is the production of Ghibli-style images from actual photographs. Users simply upload their chosen image and instruct ChatGPT in natural language to create a Ghibli-style rendition. This trend has sparked renewed concerns regarding the legality of utilizing copyrighted materials as training data for AI. It has also brought back the response from Studio Ghibli’s founder Hayao Miyazaki, who remarked years ago upon witnessing AI-generated animation, «I strongly feel that this is an insult to life itself.»

In a tweet, Altman mentioned that the tool may still show unpredictable behavior and might decline some requests that should be processed. OpenAI is actively working on addressing this issue. The company also informed TechCrunch that it «takes action» if any generated image contravenes company policies, although specifics regarding these actions were not disclosed.