Automatically detecting unauthorized re-identification

A non-transitory computer-readable medium storing instructions that, upon execution by one or more computing devices, cause the one or more computing devices to perform operations comprising:

• identifying a set of website content previously accessed by a client device;
• identifying a set of third-party content that was provided to the client device by a third-party content provider after the client device accessed the set of website content, wherein the set of third-party content differs from the set of website content;
• determining, by executing a machine learning model, a likelihood that the third-party content provider is tracking activity of the client device based on a similarity between one or more characteristics of the set of website content previously accessed by the client device and at least one characteristic of the set of third-party content provided to the client device after accessing the set of website content, wherein the determined likelihood increases as the similarity increases;
• determining, by executing the machine learning model, that the third-party content provider is tracking the activity of the client device when the determined likelihood is above a specified threshold likelihood;
• and
• causing further third-party content provided by the third-party content provider to be blocked form presentation at the client device based on the determination that the third-party content provider is tracking the activity of the client device.

A system, comprising:

• one or more storage devices storing instructions;
• one or more computing devices configured to interact with the one or more storage devices and execute the instructions, wherein execution of the instructions cause the one or more computing devices to perform operations comprising:
• identifying a set of website content previously accessed by a client device;
• identifying a set of third-party content that was provided to the client device by a third-party content provider after the client device accessed the set of website content, wherein the set of third-party content differs from the set of website content;
• determining, by executing a machine learning model, a likelihood that the third-party content provider is tracking activity of the client device based on a similarity between one or more characteristics of the set of website content previously accessed by the client device and at least one characteristic of the set of third-party content provided to the client device after accessing the set of website content, wherein the determined likelihood increases as the similarity increases;
• determining, by executing the machine learning model, that the third-party content provider is tracking the activity of the client device when the determined likelihood is above a specified threshold likelihood;
• and
• causing further third-party content provided by the third-party content provider to be blocked form presentation at the client device based on the determination that the third-party content provider is tracking the activity of the client device.