Artists Expose OpenAI’s Sora Video Generator: Defiance Against ‘AI Overlords’

During a brief period on Tuesday, a group of disgruntled artists shared a tool that granted access to OpenAI’s unreleased Sora AI model. This model, which transforms text prompts into videos, was made available to anyone temporarily.

In an open letter titled “Dear Corporate AI Overlords,” accompanied by illustrations depicting individuals gesturing with their middle fingers, the artists expressed their displeasure. They claimed to have been promised early access to Sora in order to test the product and collaborate creatively, only to discover that they were being exploited by OpenAI for their labor without compensation.

The artists outlined their grievances on the AI model-hosting platform Hugging Face, stating that while hundreds of artists provided unpaid services such as bug testing and feedback, only a select few would receive minimal compensation for their work. The majority of contributors would not be appropriately rewarded for their efforts, despite the substantial value OpenAI would gain from their contributions.

Although the artists acknowledged the potential of AI as a creative tool, they objected to the manner in which the early access program was conducted, criticizing it as a mere publicity stunt rather than a genuine opportunity for artistic exploration. They highlighted the restrictive approval process imposed by OpenAI on any videos created using the tool.

Following the uproar, the tool on Hugging Face has been disabled temporarily, and a message appended to the letter indicates that OpenAI has suspended the Sora early access program for artists.

Despite OpenAI’s initial buzz surrounding Sora, the model remains inaccessible to the public, with only a limited group of testers having experienced its capabilities. The artist group urged their peers to embrace open-source video generation tools and called on AI companies to prioritize genuine artistic expression and fair compensation for artists.