Personally identifiable information has been found in DataComp CommonPool, one of the largest open-source data sets used to train image generation models.

How funny is AI, really? Not all senses of humor are made equal. ( Undark) + What happened when 20 comedians got AI to write their routines. ( MIT Technology Review) 10 Work has begun on the first ...

AI companies generally defend their practices by claiming fair use, arguing that training AI on large data sets fundamentally changes the original content and is necessary for innovation.

Tremendous amounts of data are needed to train large language models powering generative AI. Musicians, book authors, visual artists and news publications have sued various AI companies that used ...

A federal judge has sided with Anthropic in a major copyright ruling, declaring that artificial intelligence developers can train models using published books without authors’ consent. The ...

In 2023, X changed its privacy policy to use public data on its site to train AI models. Last October, it made further changes to allow third parties to train their models .

Meta has cited legitimate interest under EU privacy rules for using users' data to train and develop its generative AI models and other AI tools that can be shared with third parties.