The maker of ChatGPT, OpenAI, sued for violating the privacy and copyright of practically everyone

The maker of ChatGPT, OpenAI, is facing a class action lawsuit alleging that the company's AI training procedures infringed the privacy and copyright of virtually everyone who has ever uploaded information online.

OpenAI gathered a sizable amount of data from diverse online sources to train its cutting-edge AI language models. These datasets include various content, including explicit content from specialized genres and Wikipedia articles, best-selling books, social media posts, and popular publications. More significantly, OpenAI collected all this data without the content producers' first consent.

In the class action case, which has been filed in California, it is claimed that OpenAI's disregard for legal protocol, including requesting permission from content providers, amounts to flagrant data theft.

"Instead of following established procedures for acquiring and using personal information, the Defendants resorted to theft," the lawsuit's filing reads. In addition to "books, articles, websites, and posts," they methodically scraped 300 billion words from the internet, which also contained personally identifiable information gathered against one's will.

If you have been active online recently, there is a good chance that OpenAI's databases contain your digital contributions. As a result, any output produced by OpenAI's language models utilized for commercial purposes might include bits of your data gathered through silent scraping.

The Washington Post was informed by Ryan Clarkson, managing partner at the legal firm suing OpenAI, that "all of that information is being taken at scale" without being initially intended for use by a substantial language model.

Nevertheless, the court case's conclusion is still up in the air. The concept of a free and open web is frequently inaccurate due to the complexity of the internet's infrastructure. Online platforms have user policies and agreements; even if people submit content to these platforms, the platform often retains ownership of the content rather than the contributors.

Intellectual property attorney Katherine Gardner pointed out that whenever users post content to social media or any other website, they typically give the platform a broad license to use that content however they see fit. Because of this, it would be difficult for regular users to assert a right to payment or other forms of remuneration for using their data in training models.