'Hard To Believe It’s AI': What OpenAI Chief Said About New ChatGPT Image Feature

Last Updated:March 27, 2025, 09:10 IST

OpenAI is finally bringing better image tools that lets people generated AI-powered images using simple prompts for high-quality results.

OpenAI is making more upgrades to the ChatGPT 4o version

OpenAI has significantly enhanced its ChatGPT chatbot, expanding its capabilities beyond text-based interactions to include advanced image generation. The company’s CEO Sam Altman described the launch as an “incredible technology/product," highlighting how it empowers users with greater creative freedom.

“I remember seeing some of the first images generated by this model and finding it hard to believe they were made by AI," Altman shared in a post on X (formerly Twitter). He emphasised OpenAI’s commitment to ensuring the technology remains within responsible boundaries while supporting a vast range of creative expression.

OpenAI released the most recent version earlier this week, which incorporates new technology that allows ChatGPT to generate intricate and realistic images in response to user input.

With the help of OpenAI’s GPT-4o model, the latest version is able to produce complex images, such as multi-panel comic strips with unique characters and dialogue. This is a significant improvement over earlier iterations that could generate images but had trouble combining several intricate ideas into a single result.

The expansion is part of a larger trend in the AI arena where chatbots are becoming more versatile tools. ChatGPT was initially created for text-based tasks like writing poetry, asking questions and creating code. Today, it may also be used for voice commands, pictures and even video interactions.

ChatGPT was first introduced by OpenAI in late 2022 after being trained on a large volume of text data from the internet. OpenAI later launched DALL-E, a separate program made specifically for image generation, even if the chatbot didn’t have that capability at first. However, the most recent version of ChatGPT combines these features into a single, more potent system that can learn from both visual and textual data.

Artificial intelligence image generators have historically had trouble producing images that differ greatly from preexisting examples. The updated ChatGPT, according to OpenAI, can now more successfully respond to requests for unusual visuals, such as a bicycle with triangle wheels.

Both free and paid users have been able to access the updated ChatGPT since March 25. The most recent AI tools from OpenAI will also be available to ChatGPT Plus ($20 per month) and ChatGPT Pro ($200 per month) subscribers.

News tech 'Hard To Believe It’s AI': What OpenAI Chief Said About New ChatGPT Image Feature