Google said it has the right to use almost any online postings to train its chatbot, Bard.

Google said it has the right to use almost any online postings to train its chatbot, Bard.  The company recently revised its privacy policy to clarify that its AI services, including Bard and Cloud AI, may use publicly available web data for training purposes. This means that any online comments or statements, regardless of their relation to Google services, could potentially be used to train AI chatbots and similar systems. The update aims to offer clarity on the company's long-standing practice of leveraging publicly accessible data for training language models like Google Translate. It now states that newer services like Bard, its AI chatbot, are included. The policy also specifies that the data will be used for "language models" rather than "AI models." The move follows the recent filing of a class-action lawsuit against OpenAI, which alleges unauthorized scraping of data for training its AI systems. Twitter and Reddit have also announced API pricing changes to limit data harvesting for AI training, though these changes have faced backlash from their respective communities.