Hollywood rejects the Seedance 2.0 model and asks ByteDance for safeguards that prevent the use of protected content

Hollywood has shared his rejection by him Seedance 2.0 artificial intelligence model, due to the ease with which it allows you to generate videos with realistic images, even starring well-known actors, without limits on the use of content protected by copyright.

Director Ruairi Robinson (‘The Silent City’, ‘BlinkyTM’) has shared a video generated with this new ByteDance tool, in which Tom Cruise and Brad Pitt fight on top of a ruined building, with the level of realism of any scene in a film shot.

It is just one example of the fidelity of the result that Seedance 2.0 offers and that has led to representatives of the film industry in the United States to report its use of protected content without permission.

This has been done by the Motion Picture Association (MPA), Charles Rivkin, of which the studios Netflix, Paramount Pictures, Prime Video and Amazon MGM Studios, Sony Pictures, Universal Studios, The Walt Disney Studios and Warner Bros. Discovery are part.

MPA President and CEO Charles Rivkin said in a statement that “by launching a service that operates without significant safeguards against infringement, ByteDance is violating established copyright legislation that protects the rights of creators and supports millions of jobs in the United States. This has led it to ask the Chinese company to “immediately cease its infringing activity.”

Disney has even sent ByteDance a cease and desist letter, in which it denounces the use of its works without permission or compensation for the development and training of the model, according to a copy of the letter obtained by Axios.

The American actors union SAG-AFTRA, for its part, has condemned “the flagrant infringement” that ByteDance makes with the new model, which includes the unauthorized use of the voices and images of its members. “This is unacceptable and undermines the ability of human talent to earn a living. Seedance 2.0 ignores the law, ethics, industry standards and basic principles of consent.”

Seedance 2.0 was presented last week as a multimodal image and video generation model which improves the quality compared to its predecessor, to obtain professional results, with realistic visual effects and dual-channel stereo sound.

Allows you to generate videos up to 15 seconds in which elements such as composition, action, camera movement, special effects and sound can be controlled in the input materials, and which offers complex movements and interactions that are synchronized and precise by realistically reproducing the laws of physics.

The appearance at the end of September last year of Sora 2, from OpenAI, generated a similar reaction from Hollywood, also due to its ability to generate videos starring characters fiction like SpongeBob, Super Mario or Pikachu and by flesh and blood actorswho did not see it right that their image and voice were replicated without their consent.