Google changed its privacy policy: Does the tech giant now use all your data to train its AI?

Google trains AI with your data

It's all in the open now: Google recently updated its privacy policy with a new paragraph:

"Google uses information to improve our services and to develop new products, features and technologies that benefit our users and the public. For example, we use publicly available information to help train Google's AI models and build products and features like Google Translate, Bard and Cloud AI capabilities."

In other words, all publicly accessible information will be used by Google to train its own AI models. This is not big news as most Big Tech companies carry out the same practices to benefit of the current AI hype, but up until now Google did not publicly include such a statement in its privacy policy.

Does private data also train AI?

The question that many people would like to see answered is whether also private data trains AI models, for instance whether your emails in Gmail train Bard. While the chatbot itself claimed it was trained with Gmail emails, Google was quick to reply, saying: "Bard is an early experiment based on Large Language Models and will make mistakes. It is not trained on Gmail data."

It is hard to know what is going on "behind the scenes" - after all Google is also saying that it is not using Gmail data to post targeted ads, which leaves people wondering why they start seeing hotel ads after having received a booking confirmation for a flight via Gmail.

It is an undeniable fact that Gmail's business model is to post targeted ads, and ad-revenue must be increased - and is being increased as you can see here - each year.

Does Google own the internet?

The new practice by Google - but also by other AI companies like Microsoft's OpenAI - means that everything that you make available online may end up in a chatbot like Bard or ChatGPT. These bots may show your content, or versions of it, as answers to different requests - without showing the source of the information.

While Google's search is not much different, it at least also shows the source. People will not only get answers to their questions, but they will also see who created the answer. They can click the link, check the website of the search result and decide for themselves whether the source is trustworthy or not.

Chatbots like ChatGPT and Bard take away the option for this second step. They decide what information to provide to whom, which gives these tools a lot of power on how to form the public opinion.

Remember AMP?

This all feels a bit like when Google introduced AMP to "make your emails faster". The aim of AMP was the same as it is now with AI trained chatbots like Bard: Google wants you to stay on their system for as long as possible. Google does not like it if you click outbound links.

The reason is obvious. As soon as you click an outbound link, Google can no longer show ads, which reduces their ad-generated revenue.

So like AMP, Bard and other AI tools are an attempt by Google to colonize freely accessible information and use it for their own benefit. Google does not want to make your search experience better or your mailbox faster, but they want to control your online usage every step of the way.

Conclusion

The best option if you do not want your data scraped, yourself being profiled and bombarded with targeted ads is to choose privacy. Today, there are a lot of privacy-first alternatives available online.

By choosing privacy, you can get away from Big Tech and their abuse of your data. Remember: Free is not free on the internet!

Fortunately, there are great alternatives to Gmail as well as to other Google products available today.

Check out our recommended, privacy-first alternatives that enable you to leave Google.