OpenAI Expands DALL-E 3 Text-to-Image Generator to ChatGPT

New Image Generator, DALL-E 3, Available to Paying Customers, Can Generate More Detailed Imagery, Safeguards in Place For Ethical Usage, and Working on Image Classifier for AI-Produced Disinformation

OpenAI has expanded availability of its latest text-to-image generator, DALL-E 3, to paying customers of its ChatGPT Plus and Enterprise chatbot services. The image generator can be utilized within the ChatGPT app and can create more detailed images based on intricate, detailed prompts created by ChatGPT.
DALL-E 3, an improvement on its predecessor DALL-E 2, has the ability to take in better context and therefore generate more precise and visually striking images. It can also handle a range of different aspect ratios and can render intricate details and text. Microsoft's Bing Chat was the first chatbot to integrate with DALL-E 3, before ChatGPT.
In order to prevent unethical usage, OpenAI has put in place various safeguards, such as not allowing DALL-E 3 to create images of public figures. However, loopholes still exist as evidenced by Microsoft's integration, where users were able to churn out inappropriate images using the DALL-E 3 image generator.
OpenAI is now working on an image classifier tool which will be able to detect, with up to 99% accuracy, whether an image was generated by DALL-E 3. This tool will be a valuable weapon in the fight against AI-generated disinformation and propaganda.
Despite the introduction of robust safety measures, there has been some controversy with misuse of DALL-E 3's capabilities, such as generating offensive imagery or contributing to disinformation campaigns. OpenAI has pledged to take additional steps to mitigate these issues.