During a livestream on Tuesday, OpenAI CEO Sam Altman unveiled the first significant upgrade to ChatGPT’s image-generation capabilities in over a year.

Figure 1. Enhanced Image Generation with ChatGPT.

ChatGPT can now utilize the company’s GPT-4o model to natively create and modify images and photos. While GPT-4o has powered the AI chatbot platform for some time, it was previously limited to generating and editing text—until now. Figure 1 shows Enhanced Image Generation with ChatGPT.

Altman announced that GPT-4o's native image generation is now live in ChatGPT and Sora, OpenAI’s AI video-generation tool, for subscribers of the $200-a-month Pro plan. OpenAI also stated that the feature will soon roll out to Plus and free ChatGPT users, as well as developers using the company’s API service.

GPT-4o with image output takes slightly longer to process compared to the model it replaces, DALL-E 3, but OpenAI claims it produces more accurate and detailed images. The model can also edit existing images, including those featuring people, by transforming elements or "inpainting" details such as foreground and background objects.

To support the new image-generation feature, OpenAI told The Wall Street Journal that it trained GPT-4o using “publicly available data” and proprietary data from partnerships with companies like Shutterstock.

Many generative AI companies treat training data as a competitive asset, keeping details tightly guarded. Additionally, concerns over intellectual property disputes provide another reason for companies to disclose little about their data sources.

OpenAI provides an opt-out form that allows creators to request the removal of their works from its training datasets. The company also states that it honors requests to block its web-scraping bots from collecting training data, including images, from websites.

ChatGPT’s upgraded image-generation feature comes shortly after Google’s experimental native image output for Gemini 2.0 Flash, one of its flagship models. While the feature gained viral attention on social media, it also faced criticism due to weak safeguards, allowing users to remove watermarks and generate images of copyrighted characters.

The Big Announcement

OpenAI CEO Sam Altman recently announced a major enhancement to ChatGPT’s image-generation capabilities, marking the first significant upgrade in over a year. This improvement is powered by the company’s GPT-4o model, which can now natively generate and modify images alongside its existing text capabilities.

How GPT-4o Enhances Image Generation

GPT-4o replaces DALL-E 3 as ChatGPT’s core image-generation model. While it takes slightly longer to generate results, OpenAI claims it produces more accurate and detailed images. Additionally, GPT-4o introduces advanced editing features, including the ability to modify existing images and “inpaint” missing or altered details.

Access and Availability

The upgraded image-generation feature is currently available to Pro plan subscribers ($200/month) within ChatGPT and OpenAI’s AI video tool, Sora. However, OpenAI has confirmed plans to roll out the feature soon to Plus and free-tier ChatGPT users, as well as developers using the API.

Data Sources and Ethical Considerations

OpenAI states that GPT-4o was trained using publicly available data and proprietary sources from partnerships, such as Shutterstock. To address copyright concerns, OpenAI offers an opt-out form for creators and respects requests to block its web-scraping bots from collecting training data.

The Competitive Landscape and Industry Challenges

ChatGPT’s image upgrade follows Google’s Gemini 2.0 Flash, which recently made headlines for its controversial lack of guardrails, allowing users to remove watermarks and generate copyrighted images. As AI-generated media continues to evolve, companies face growing scrutiny over intellectual property rights, ethical AI use, and responsible content generation.

Source:TC

Cite this article:

Priyadharshini S (2025),”ChatGPT's Image Generation Feature Receives an Upgrade", AnaTechmaz, pp. 230