OpenAI CEO Sam Altman announced the rollout of image generation in ChatGPT, as the company shared a sample image that highlighted how accurate the text was when compared to older AI-generated images.

Image generation was previously possible through the DALL-E models, but the results from these models were highly distorted, contained many mistakes, and were easy to flag as being AI-generated.

OpenAI has now brought image generation to ChatGPT itself, or GPT4o, to be specific.

The AI company backed by Microsoft shared one promotional image that it said was generated with 4o, where it pointed out the accurate, distortion-free text on a picture of a boarding pass.

“Creating and customizing images is as simple as chatting using GPT‑4o—just describe what you need, including any specifics like aspect ratio, exact colors using hex codes, or a transparent background,” posted OpenAI on X on March 26.

The new image generation feature is set to roll out in ChatGPT and Sora to all Plus, Pro, Team, and Free users.

Altman hailed the company’s “new high-water mark” in allowing the creative freedom offered by the new feature. However, he warned that some users might be able to create content that offends others.

“[A]s we talk about in our model spec, we think putting this intellectual freedom and control in the hands of users is the right thing to do, but we will observe how it goes and listen to society. we think respecting the very wide bounds society will eventually choose to set for AI is the right thing to do, and increasingly important as we get closer to AGI. thanks in advance for the understanding as we work through this,” posted Altman on X on March 25.

In response to a user who claimed that they could general sexual pictures of men but not women because of the platform’s content filters, Altman responded this was a bug that would be fixed.