Chrome 101 ships with controversial ad system tests
Google plans to release a new stable version of the company’s Chrome web browser later today. Chrome 101 is released four weeks after the release of the last version of Chrome, which was Chrome 100.
The new version is not particularly interesting in terms of the new features that Google introduces in it. The Chrome Status Page lists only a few, and most of the features listed on the page are of interest only to developers.
Controversial tests of advertising APIs begin
Two features that may be of interest to many Chrome users are not listed on the page. They may not be listed because they run as tests in the browser. So-called source testing introduces features to Chrome for a subset of users, often to give developers access to these features to deploy and test services.
The two functions, the Theme API and the first Locally Executed Group Decision Experiment (FLEDGE), run as source tests on Chrome 101 through 104 on the desktop.
To better understand what they do, it is necessary to look back at the development of Google’s privacy sandbox initiative and the intention to remove third-party cookies on the Internet. Google’s main source of revenue comes from advertising, and the removal of third-party cookies poses a threat to its core business. Google is now in a position to find something equally lucrative and at the same time less intrusive when it comes to user privacy on the Internet.
Google’s first attempt to establish a replacement was called FLoC. Announced several years ago, FLoC, or Federated Learning Cohorts, moved tracking from individual users to groups of users. Many companies and organizations, including Brave, DuckDuckGo, and Vivaldi, criticized FLoC and announced that they would block it in their browsers and products.
Criticism focused on several fundamental objections:
- With FLoC in place, sites would be informed about the interests of a user’s group, even if the particular user has never visited the site.
- FLoC adds another bit of data to the fingerprint. Since an FLoC is made up of a few thousand users, it is a small group and as such ideal for fingerprinting.
- Google is in control of sensitive topics, which are not included in the interests disclosed to sites and advertisers.
Google removed FLoC in early 2022 and announced that it would use Themes instead, for its future cookieless advertising system. Themes run locally in the web browser. The API uses algorithms to determine topics of interest based on the user’s browsing history. Interests are stored locally in the browser for three weeks. When a user visits a site, three of the interests of the site in question and its advertising partners are revealed.
PLUMAR, the first “Locally Executed Decision on Groups” experiment, is the second advertising technique that Google tests in Chrome versions 101 to 104. The technique moves “the data of interest and the final decision on the ad” to the browser local. Google hopes the technique addresses major privacy concerns while giving advertisers enough data to show ads that may be of interest to users.
Topics and FLEDGE are not exempt from criticism either. Both don’t address the issue with sensitive topic selection. A Microsoft Edge developer published a short script in mid-2021 that could be used for cross-site tracking with FLEDGE.
Chrome 101 will be released later today.
Now you: What is your opinion about these new techniques?
advertising
