Google is set to release its latest targeted-advertising API, Topics, with the launch of Chrome 115 on July 12. This new mechanism will enable advertisers to deliver personalized ads while prioritizing privacy.
Google has explicitly requested advertisers not to misuse the API and to handle user data responsibly.
The request came about after Criteo’s senior data scientist and software engineer, Alexandre Gilotte, identified the potential for a fingerprinting attack through the Topics API. He hinted at the possibility during a GitHub issues discussion last year.
The Controversy Over Topics
Notably, the Topics API functions by enabling the delivery of ads based on inferred user interests. It fetches a list of up to three topics from a taxonomy of several hundred interest categories derived from the user’s past website visits.
This allows the site to show an ad that aligns with the visitor’s known interests. The API also includes a feature that may return a random topic.
Vinay Goel, product director of Privacy Sandbox at Google, elaborated on the operation of Topics in 2022, clarifying that topics are kept only for three weeks and are selected entirely on the device without involving external servers.
Subsequently, they may develop a fingerprint identifier.
Firefox maker Mozilla and Safari developer Apple expressed skepticism towards Google’s measures. According to Mozilla engineer Martin Thomson, Topics somehow fails to provide meaningful protection for privacy.
We don’t think cross-site data about the user’s browsing behavior should be exposed in APIs.Anne van Kesteren, Apple
Last week, Google’s privacy Sandbox project manager, Josh Karlin, announced that the company had added a Chrome requirement asking developers to attest that they won’t abuse the API.
Google’s Response to the Concerns
Responding to these concerns, Google contended that Topics offers better privacy than third-party cookies.
The company released a paper in April to evaluate that claim. Additionally, the ad giant recently introduced changes to the Topics API. The said modifications include a new taxonomy of 469 interest topics and the capability for users to block specific topics.
We chose to limit the taxonomy’s size to protect against re-identification risk.Leeron Israel, product manager for Google’s Privacy Sandbox
However, Mozilla is still not convinced about the efficacy of these modifications. A spokesperson from Mozilla stated that Google is content to utilize down-market measures to bring in a sense of privacy.
Such measures won’t console users who are reidentified using that information.
Nevertheless, despite the blazing controversies, Google has chosen to move forward with Topics. The company abandoned its Federated Learning of Cohorts (FLoC) last year, expressing its interest in facilitating interest-based advertising.
While Topics’ user-friendliness and usefulness suggest a bright future for interest-oriented advertising, the concerns around the API’s privacy protection feature are somehow suppressing its acceptability and reliability. Whether the API will manage to get over these concerns remains an open question.