A class-action lawsuit filed in California alleges that OpenAI and Microsoft illegally used personal data from millions of internet users to train their AI technologies, including ChatGPT. The lawsuit claims the companies "scraped" 300 billion words from the internet without consent, violating privacy and copyright laws. This "unprecedented theft," the lawsuit argues, is the foundation of their multi-billion dollar businesses. The plaintiffs contend that the AI products were rushed to market without adequate safeguards, posing real dangers. The suit quotes OpenAI CEO Sam Altman's statement about AI's potential world-ending capabilities, though acknowledging it may have been a joke. The lawsuit seeks to halt commercial access to and development of OpenAI's products, establish an oversight council for AI product approval, and provide compensation to users for their stolen data. OpenAI updated its data policies in March, stating it wouldn't use API-submitted data for training without explicit consent. Data is retained for 30 days for misuse monitoring, then deleted unless legally required. OpenAI emphasizes that data is used to enhance model helpfulness, not for sales, advertising, or profiling. Microsoft, a major investor in OpenAI, declined to comment. While other companies use internet data for AI training, the lawsuit targets OpenAI as the "natural first target" for sparking an "AI arms race."

OpenAI with Microsoft Bing on mobile.

Sam Altman, CEO of OpenAI.

Microsoft Campus in Redmond, Wash.