Artificial Intelligence (AI) is here to stay and slowly but steadily increasing its influence on our (on line) lives. This is a blessing in disguise: AI (and its cousin Machine Learning) opens up endless possibilities and will have a profound impact on how we deal with the increasingly digital world surrounding us. It will also surrender us to incomprehensible algorithms, compare us to peers we did not even know existed and profile us to the extreme. There will be machines that know us better than we do.
Is this a bad thing? I guess we do not know. It does raise a question, however. Are we able to outwit intelligence? Just for the sake of it, if nothing else. In this blog series, I will feature people and projects that are cautious and curious about AI. As the technology develops, we have an amazing opportunity to study, learn and possibly even influence it to some extent. The strategy to confuse AI is called obfuscation, and it yields interesting results and examples.
Part 1: AdNauseam
In the first part of the series, the open source controversial browser extension AdNauseam takes the stage. It works by automatically and randomly clicking on advertisement links on any webpage you visit. User tracking and specific ad targeting becomes quite impossible. Needless to say, this does not sit well with large ad networks such as run by Google.
Why
While AdNauseam might sound like a playful extension to an ad blocker, there is some serious research behind it. Thought up by Hellen Nissenbaum, Daniel C. Howe and Mushon Zer-Aviv in 2009, the theory behind the tool is described in an academic paper dating back to 2017 and published in the proceedings of the 3rd International Workshop on Privacy Engineering: Engineering Privacy and Protest, a Case Study of AdNauseam. Quite simply, AdNauseam was designed to confuse advertisement networks, in order to protect users from tracking.
What does it do
AdNauseam is built on top of the browser extension uBlock and in its simplest form blocks or hides ads. Next to that, it offers users the ability to look into the ads that appeared on web pages they visited. By doing so, it helps you understand why certain ads were served to you on specific websites – providing you with an insight into the algorithms driving the ad networks. Finally, the tool actually clicks on these ads randomly, in order to obfuscate the ad network trackers.
Goals
AdNauseam is driven by more than just the desire to block ads and protect from potentially harmful advertising. It is an instrument that allows users to signal to the large advertising networks that they understand how advertising on the internet works, and that they do not necessarily agree with that. As stated by Nissenbaum, “Rather than enacting privacy as concealment, AdNauseam provides a means for users to express, in plain sight, their dissent by disrupting the dominant model of commercial surveillance.”
General reception and criticism
In 2017, Google removed AdNauseam from the Chrome Web Store, stating it violated the following part of its terms of service: ““an extension should have a single purpose that is clear to
users.“ AdNauseam was also labelled as malware. Google probably meant to say “an extension should not violate our business model”. The business model discussion is a recurring one – it is about advertisements driving most of the income generated on the internet, which in turn does benefit (sm)all website owners.
But, as Nissenbaum states, “Whether or not one is an advocate of obfuscation, it is disconcerting to know that Google can make a privacy extension, along with stored data and preferences, disappear without warning.“
In 2019, Nissenbaum spoke at Cornell Tech about the power of obfuscation. Audience members challenged the way AdNauseam worked and stated that the technology behind advertising networks could and would easily recognize the tool. This led to an increased effort to prove that AdNauseam works, but also led to more research on how advertising networks such as Google actually work. The results are described in more detail in an article on MIT’s Technology Review.
More controversy
In 2017, privacy focused browser Pale Moon blocked the browser extension as well, stating that AdNauseam does not necessarily constitute click fraud, but does harm the internet economy as a whole. This led to a heated debate between the developer and some of its users, including on forums such as Hacker News. Even though this is old news, the discussion remains. It is not really a question anymore on the effectiveness of AdNauseam, but more about whether guerrilla tactics are allowed to disrupt a business model that ultimately harms our privacy, but yields income to a large part of the internet – not just Google.
Alternatives
TrackMeNot is another browser extension created by Nissenbaum and Howe, which runs randomized search queries in the background, aiming to obfuscate user profiling by search engines.
A number of advertisement blockers exist, which offer different levels of sophistication in, well, blocking ads. A tool that goes a step further is Privacy Badger by the Electronic Frontier Foundation: it works invisibly in the background, blocking (cookies from) advertisement trackers automatically after it has seen them appear on three different websites. Less guerrilla, similar outcome?