OpenAI vouches stronger guardrails against deepfake generation via Sora

OpenAI has said that it will increase oversight over its video generation app Sora to prevent it from being exploited for deepfakes. The development comes after “Breaking Bad” actor Bryan Cranston raised concerns about his identity being misused in unauthorized AI-generated content on the platform.

Cranston reached out to actor unions including SAG-AFTRA to express his issues after which OpenAI reportedly promised to put stronger guradrails in place to keep Sora from becoming a hotspot of deepfake generation.

According to a statement by SAG-AFTRA, OpenAI’s opt-in policy for uses of an individual’s voice or likeness in Sora 2 is to give all artists, performers, and individuals the right to determine how and whether they can be simulated.

“This policy reflects the company’s commitment to artist rights, transparency, and the ethical deployment of generative technology. It has also committed to responding expeditiously to any complaints it may receive,” the post by SAG-AFTRA said on X.

The post also added that it supports the formation of a No Fakes Act to establish a national standard that ensures that the identities of public figures are not exploited to spread fake information.

“I am grateful to OpenAI for its policy and for improving its guardrails, and hope that they and all of the companies involved in this work, respect our personal and professional right to manage replication of our voice and likeness,” CNBC quoted Cranston as commenting on the development.

OpenAI has now recently stuck a deal with Mattel to test Sora.

Earlier this month, OpenAI wrapped up a secondary share sale that saw $6.6 billion worth of stock sold as the firm achieves a valuation of $500 billion, making the company the biggest startup.