Sora Leak: OpenAI Accused of Artist Exploitation | OpenAI free | OpenAI italiano | Chat GPT login free | Turtles AI

Sora Leak: OpenAI Accused of Artist Exploitation
Activists reveal early access to OpenAI video generator, denounce unfair practices and lack of compensation for creators involved
Editorial Team27 November 2024

 

A group of activists has published a project that provides access to OpenAI’s Sora video generator, which is not yet publicly available. The group, called “Sora PR Puppets,” denounces the misuse of artists as unpaid testers and a lack of transparency in the development of the tool. OpenAI revoked access shortly after the leak.

Key points:

  • The group "Sora PR Puppets" has released a frontend that allows you to generate videos with Sora, via early access. 
  • The initiative was born as a protest against the practices of OpenAI, accused of exploiting artists without compensation. 
  • OpenAI temporarily disabled early access after the system was made public. 
  • Sora, OpenAI’s video generator, continues to encounter technical difficulties and suffer from competition from other players in the sector.

OpenAI’s Sora video generator, a highly anticipated project in the AI ​​space, recently gained attention for a controversial leak. An anonymous group known as “Sora PR Puppets” posted a project on the AI ​​development platform Hugging Face that allows users to generate short videos using early access to the Sora system. Although the Sora API was not yet publicly available, the group created an interface that, through a simple text prompt, allows for approximately 10-second videos at up to 1080p resolution. The project has piqued the interest of many, including TechCrunch, which reported long lines to use the service. However, shortly after the news broke, access to the system was disabled by OpenAI and/or Hugging Face, quickly ending the parenthesis.

The motivation behind the group’s action appears to be a protest against what they call deceptive and unethical practices by OpenAI. According to the activists, the company is exploiting the free work of many artists involved in Sora’s early access. These artists, including the so-called "red teamers" and creative partners, are said to have been pushed to promote the video generator, creating a positive narrative around the product without receiving adequate compensation. The group also denounces the lack of freedom for creators, who would not be able to freely share the material produced, since each output generated must be approved by OpenAI before being released.

Furthermore, the activists criticize the company for the opaque management of the early access program, accusing it of favoring public relations more than real artistic and creative development. The group, while not opposed to the use of AI technology in the field of the arts, disputes the way in which the program has been structured and the perception that is emerging. For its part, OpenAI responded that participation in the early access program is voluntary and that the company has supported artists with grants and support programs, reiterating that access remains a phase of research and development.

Sora launched this year, but has faced numerous technical hurdles. Early versions of the system were slow to process, with one-minute videos taking over 10 minutes to complete. The results were also inconsistent, with difficulty maintaining consistent style and objects across the video. Recently, there has been talk of a “turbo” version of the system, which would be faster, but with limited customization options. Despite the challenges, OpenAI continues to invest in improving Sora, hoping to outpace the increasingly fierce competition, as evidenced by moves by rival companies such as Runway and Stability AI, which are partnering with major movie studios to develop custom video generation models.

Amid these challenges, OpenAI has yet to sign any significant partnerships with major movie studios, despite meeting with filmmakers earlier this year.

While it is clear that the Sora project represents an ambitious attempt to enter the AI ​​video generation sector, the company will have to face numerous technical and image obstacles to consolidate itself as a reference player.