Google Updates Privacy Policy to Utilize User Data for AI Development

Artificial intelligence (AI) models are trained using data sets that enable them to analyze and respond to texts in various contexts and languages. For example, ChatGPT has undergone training using a large text dataset that is publicly available. In contrast, DarKBERT is an LLM that has been trained on an extensive dataset consisting of dark web pages, gathering information from sources like hacker forums, scamming websites, and other criminal internet platforms. Due to the insatiable appetite of AI tools for data, any content posted online by individuals is considered fair game. Google has recently updated its privacy policy, clarifying that any online posts can now be utilized to train its AI tools and models.

New privacy policies

Google announced changes to its privacy policy on its website. It states: “Google uses data to improve its services and develop new products, features and technologies that benefit our users and the public. We use publicly available data, for example, to train Google’s AI models and build products and features such as Google Translate, Bard and Cloud AI – features.

A review of Google’s privacy policy history tells about the changes made by Google. Google previously stated that your data may be used in “language models”, but this has now been replaced by “AI models”. Additionally, the practice previously only mentioned Google Translate, but Cloud AI and Google Bard are now included.

While most companies’ privacy policies include the right to use any data published on their platforms, Google now reserves the right to collect and use data posted online as a whole, which is used to develop its services and train AI models.

How do artificial intelligence models get their information?

Generative AI models like ChatGPT use the entire Internet to obtain their data through a process called Web scraping. It extracts a valuable amount of information from online sources and then provides an opinion analysis of it to the user. While web scraping may be useful for analytical research purposes, it may also violate the terms of service of a website that prohibits web scraping.

To combat the extreme level of data scraping and system manipulation, Elon Musk recently restricted Twitter accounts to a limited number of reads per day. In addition, Twitter also limited users’ browsing rights without an account. Read more here: Elon Musk will change Twitter forever, limit the number of tweets that can be read.