OpenAI CTO Mira Murati faces backlash for lack of transparency about training data for Sora.

From Futurism:

OpenAI’s CTO Mira Murati struggles to pinpoint the exact source of data used to train its new text-to-video AI, Sora. When asked about training data from YouTube, Instagram, or Facebook, Murati offers vague responses, leading to controversy over OpenAI’s data-scraping practices and lack of transparency.

Following a cringe-inducing interview with The Wall Street Journal, where Murati couldn’t specify the training data for Sora, reports later confirm that Shutterstock videos were used. Despite mixed reactions online, concerns persist about AI companies using publicly available data without clear transparency or accountability.

The lack of transparency from OpenAI raises questions about data privacy and copyright issues in the digital age. Regardless of whether Murati’s evasiveness was intentional or due to a lack of knowledge, the public demands clearer answers regarding the sources of data used in AI training. Vague corporate responses are no longer sufficient in addressing these concerns.



Read more at Futurism: In Cringe Video, OpenAI CTO Says She Doesn’t Know Where Sora’s Training Data Came From