Sora
What was Sora?
Sora was an AI text-to-video generator developed by OpenAI, the company behind other AI systems such as ChatGPT and DALL·E. Sora produced realistic videos based on textual prompts written by users. Sora means ‘sky’ in Japanese. It could generate videos up to a minute long. The footage produced by Sora could include multiple characters, different types of motion and cinematic detail in the background.
Sora ran on an adaptation of the DALL·E 3 model, which was trained using millions of images from the web. OpenAI stated that Sora also used photo and video content from the content library Shutterstock.
In 2024, OpenAI released Sora to a handful of AI experts and creative practitioners (visual artists, designers and filmmakers) to test the platform's capacities and its potential harms and risks. Among these risks was the production of abusive, shocking or problematic content. Because the videos produced by Sora were incredibly realistic, people could find it challenging to determine if a video generated by Sora was real or fake. This could lead to the indiscriminate use of deepfakes for disinformation and other harmful content.
Website: openai.com/sora
How did people use Sora?
Learn more about the benefits and risks associated with how people used AI video generation tools like Sora.
Sora was used for: ai generative images, photo/video sharing and screen capture.
Last updated: 10/09/2026