OpenAI CEO Sam Altman announced the first significant upgrade to ChatGPT’s image-generation abilities in over a year during a livestream on Tuesday.
The new feature allows ChatGPT to natively create and modify images and photos using the company’s GPT-4o model, which has long powered the AI chatbot but, until now, was limited to generating and editing text.
Altman revealed that the native image generation feature is available today for subscribers to OpenAI’s $200-a-month Pro plan, and is also live in Sora, the company’s AI video-generation product. The feature will soon roll out to Plus and free ChatGPT users, as well as developers using OpenAI’s API service.
GPT-4o’s image generation is designed to deliver more accurate and detailed results compared to the model it replaces, DALL-E 3. Although GPT-4o may take a bit longer to generate images, it provides enhanced capabilities such as editing existing images—including those with people—by transforming elements or inpainting details like foreground and background objects.
To support this new feature, OpenAI trained GPT-4o using “publicly available data,” along with proprietary data from partnerships with companies such as Shutterstock. OpenAI, like other AI companies, keeps specifics about its training data confidential, as it can be a competitive advantage and a potential source of legal challenges over intellectual property.
“We’re respectful of artists’ rights in terms of how we do the output, and we have policies in place that prevent us from generating images that directly mimic any living artists’ work,” said Brad Lightcap, OpenAI’s chief operating officer, in a statement.
OpenAI also provides an opt-out form for creators who wish to have their work removed from the training datasets. Additionally, the company respects requests from website owners to prevent its web-scraping bots from collecting training data, including images, from their sites.
This upgrade follows Google’s recent introduction of native image generation in its Gemini 2.0 Flash model, which gained significant attention on social media. However, the feature also sparked controversy due to a lack of guardrails, allowing users to remove watermarks and create images featuring copyrighted characters.
