Advertising "cookies" did not take their name from pastries... but from LSD. The original name for these bits of code that machines exchanged was "magic cookies". Today, cookies are considered too intrusive, and two browsers have already changed their approach. (Photo: Shutterstock)

Advertising "cookies" did not take their name from pastries... but from LSD. The original name for these bits of code that machines exchanged was "magic cookies". Today, cookies are considered too intrusive, and two browsers have already changed their approach. (Photo: Shutterstock)

On Tuesday, Google presented its new way of tracking the activity of Internet users. Bye bye Floc, hello Topics, whose development will have to be monitored this year to measure its real impact on the protection of privacy.

Would you want a stranger to follow you to the toilet and then to the coffee machine and record your every move? This stranger doesn’t say anything, but studies your every move, until one day they know enough to anticipate the moment you decide you want an orange juice, and offers you another orange juice instead. They want to replace you in every moment, rendering you into an object of study and a kind of laboratory rat to which private companies and armies of marketers have access to sell their products and services.

In the virtual world, the stranger is the advertising cookies, the trace we leave when we go online. They have been called "cookies" since the beginning of the internet, when developers looked at those lines of code used for identification exchanged by machines and called them "magic cookies.” This name is borrowed from the LSD register and was first used in 1994 by Lou Montulli. Although the vast majority of Internet users are generally indifferent to what goes on in the background and are often not even aware of it, debates on privacy are forcing the major operators to move.

Too much negative feedback for Floc

On Tuesday, Google announced that it was ending Floc (federated cohort learning). Roughly speaking, this idea put each internet user in a pool of internet users who share the same interests to make the individual untraceable.

The cohort was to be updated weekly and each individual could be placed in another cohort depending on their activity. Whenever a cohort connected "too much" to certain sensitive themes, such as race or sexuality, the cohort was automatically "destroyed" and the individuals were moved to another group.

Today, in light of the negative feedback from this system, which was designed to allow private companies and marketers to continue sending targeted advertising without knowing everything about an Internet user, the American giant has decided to replace it with Topics, based on a different dynamic.

350 themes out of 1,500 available

In Topics, there are no longer groups of Internet users with similar profiles, but rather a classification of the Internet user's activities according to "centres of interest". Of the 1,500 that appear in the international taxonomy, Google promises to retain only 350, which will be updated or selected by humans. In other words, not by artificial intelligence based on the aggregation of Internet users' browsing habits.

Each week, the user will be able to see the list of six themes that categorise his or her browsing and delete those that he or she does not like. And only three of the "qualified" sites for each of the six themes will be able to receive data on a user, provided that this data relates to the last three weeks. The technical documentation shows very well the thousands of tinkerings that will be required.

With 70% of the market, the giant will be the last to look for a solution to limit "third-party cookies” --this category of trackers that advertisers love to profile. Apple took care of it in 2017 for Safari, and Firefox did the same the following year.

We will have to see what results from this change, in light of the biggest failure of the European Data Protection Regulation: how the internet user can free himself from this organised gang profiling. The GDPR has failed because we click away without even looking at the window that appears, indifferent to what is happening in the background.

This story was first published in French on Paperjam. It has been translated and edited for Delano.