Sora is built on the research of DALL·E and GPT models and can animate static images and create complex scenes.
Currently, Sora is accessible only to red team members and select artists for feedback.
OpenAI plans to make Sora available to all users eventually, but safety measures are being implemented and rigorous testing is being conducted to address potential risks.
OpenAI acknowledges safety concerns about Sora's realistic video generation. They are collaborating with experts to test and implement safety measures prior to a public rollout. Detection tools will be developed to address potential risks and maintain content integrity.