Neil Richards, Woodrow Hartzog, Jordan Francis

When is surveillance advertising, the personalized marketing to people based upon their habits and traits, a privacy problem? Lawmakers and regulators have struggled to mount a coherent approach to describing the risks and benefits of ad delivery based upon pervasive electronic information collection. Before 9/11 and the second Internet boom, such a business model would have been technologically, economically, and politically impossible. Moreover, it was never disclosed or discussed - it was just developed by companies in the shadows. But it has become, for better or for worse, one of the default business models of economic commerce.

Lawmakers have struggled to regulate the excesses of surveillance advertising because it is often difficult to categorize ad tech practices as careless or unfair (though sometimes it is). Surveillance advertising’s critics have also struggled to overcome the response that it powers significant “free” services and offers “more relevant ads” that are per se better than “less relevant ads” that are not based upon surveillance and targeting. 

This paper offers such a response. Drawing on prior work by the authors and by the large interdisciplinary field of privacy scholarship, it argues that surveillance advertising’s real problem is that it is disloyal. Surveillance advertising’s benefits flow overwhelmingly to advertisers and platforms while its dangers are borne almost entirely by the vulnerable people on the other side of an information relationship with a massive power disparity. In this respect, advertising is not merely a default model of the internet economy, it embodies and illustrates some of its greatest flaws.

In this essay, we develop the concept of “disloyal advertising” as a framework for regulating ad tech and targeted marketing. Evocative of prior categories like “subliminal advertising” and “unfair and deceptive trade practices,” the notion of “disloyal advertising” goes beyond prior understandings to explain exactly why surveillance advertising is a problem and how to fix it. Inspired by classic fiduciary duties of loyalty, loyal advertising duties would successfully limit the abuses of platforms, ad tech, and data brokers while fostering sustainable marketing practices and a safe and fruitful personalized Internet.

Our essay proceeds in three parts. First, we explore how surveillance advertising is inherently disloyal because it disproportionately benefits companies at the expense of individuals and society. We draw from the literature on behavioral targeting to distinguish between standard contextual advertising systems, which require minimal data, from cross-context behavioral advertising systems, which demand troves of data and imperil people.

Second, we make the case for a duty of loyalty for surveillance advertising. We argue that a loyalty approach to surveillance advertising has three major benefits over the traditional “notice and choice” approach to privacy that prioritizes an individual’s control over their data and allows any advertising practice so long as it is not deceitful or unreasonably harmful. Loyalty rules prevent platforms from using a threadbare notion of consent to justify dangerous practices that primarily benefit platforms. Particularly when it comes to technologies like surveillance advertising that operate at or below the waterline of consumer consciousness, eople should be protected no matter what they choose. Loyalty rules also would allow regulators to avoid the problem of demonstrating that financial, emotional, or reputation harm is a “substantial injury” as part of an unfair trade practices claim. The key harm from disloyal advertising is the corruption of the relationship of trust between people and organizations. No further proof of harm is needed, though there are certainly others, including manipulation and the development of technologies that can be used in disloyal political advertising. Finally, loyalty rules add an additional factor to be considered by judges and enforcement agencies: the relative benefit of surveillance advertising to organizations. While some people might modestly benefit from “more relevant ads,” any such meager gains pale in comparison to the windfall reaped by platforms and ad tech companies. Comparing both the risks and benefits of both parties reveals that ad relevancy is a flimsy pretext to justify the largest data extraction operation the world has ever known. 

Third, we propose a legal framework for regulating disloyal advertising. Specifically, we propose leveraging the concepts of data minimization and the California Consumer Privacy Act’s notion of “cross-context behavioral advertising” to supplement a general duty of loyalty for personalization. Robust rules prohibiting platforms and ad tech companies enriching themselves at the expense of trusting parties won’t ruin the Internet or eliminate personalized technologies. But it will eliminate those that must rely upon disloyal behavior to survive.