The update, which was made over the weekend, was first discovered by Gizmodo, who reported changes to the policy’s research and development section, which details how Google can use users’ information.
With the update, the section now reads: For example, we may collect information that’s publicly available online or from other public sources to help train Google’s AI models and build products and features like Google Translate, Bard, and Cloud AI capabilities.
Prior to the update, the section only mentioned training for “language models,” particularly Google Translate. Now, the update explicitly mentions AI models, Bard (Google’s conversational generative AI chatbot), and Cloud AI.
However, some experts have raised concerns that the shift in the language in Google’s policy update may suggest that it can now use the information posted online – an issue that raises questions around privacy and intellectual property rights and whose legality is still being threshed out in courts.
Recently, ChatGPT creator OpenAI was the subject of a lawsuit claiming that the company trained its AI products using stolen personal data.
According to the complaint, OpenAI resorted to “theft” when it systematically scraped 300 billion words from the internet, which included personal information, without consent.
The lawsuit said that the data was then used to develop AI products using large language models and deep language algorithms to study and generate human-like language that can be used for a wide range of applications, including chatbots, language translation, and text generation.