Google Harnesses Public Web Data to Train AI Models

Google has updated its privacy policy to disclose that its AI services may be trained on publicly available data scraped from the web. The policy now clarifies that services like Bard and Cloud AI are included in the use of public data. However, it does not specify how copyrighted materials will be prevented from being included in the data pool. This approach raises questions about compliance with global regulations like GDPR and the fair use doctrine. The use of scraped data by AI companies has sparked lawsuits and calls for stricter regulations. Additionally, the working conditions of those sorting through the vast amounts of training data have come under scrutiny.