Everything you post online will be read by Google’s artificial intelligence

Google has changed its privacy policy to explicitly state that anything that appears on the public internet can be used by the company to develop and train AI tools.

Which means that anything you type on the internet will eventually be absorbed by the AI.

“Google uses the information to improve our services and to develop new products, features, and technologies that benefit our users and the public,” says the new company policy.

“For example, we use publicly available information to train Google AI models and develop products and features such as Google Translate, Bard, and Cloud AI.”

Gizmodo points out that the previous privacy policy only said that the data would be used for Google Translate and “language models,” rather than “AI models” in general.

It’s an odd change, since companies’ privacy policies usually refer to the use of data entered on the company’s own services, rather than anywhere on the Internet.

Data war

Until today we knew that everything we publish on the World Wide Web can be accessed by any user. But the issue now is not only who reads this content but also how it is used on an industrial scale.

Many models, such as OpenAI’s ChataGPT, are trained on copyrighted data, a practice that seems to move into a legal gray area.

Artists and companies like Getty Images have sued AI companies over the use of their intellectual property, a number of issues that still have to be heard in court.

Meanwhile, the massive data collection has caused controversy between AI creators and companies like Reddit and Twitter, which have blocked free access to their APIs, through which anyone can download large amounts of data.

Indeed, Elon Musk has accused Microsoft, which funds OpenAI, of “illegal” access to large amounts of Twitter data.

According to Musk, this was the reason Twitter imposed a temporary limit on the number of tweets one could read.