Chapter 16: Further Reading

Transparency in AI Marketing and Advertising

Annotated bibliography of 18 essential sources.

Investigative Journalism and Primary Accounts

1. Angwin, Julia, Terry Parris Jr., and Surya Mattu. "Facebook Lets Advertisers Exclude Users by Race." ProPublica, October 28, 2016. The original investigation that documented Facebook's Ethnic Affinity advertising targeting and demonstrated through test purchases that advertisers could exclude Black and Hispanic users from housing ads. This investigation triggered the enforcement actions, litigation, and regulatory attention that produced the HUD settlement. Essential reading for understanding how discriminatory advertising targeting was built into Facebook's platform.

2. Cadwalladr, Carole, and Emma Graham-Harrison. "Revealed: 50 Million Facebook Profiles Harvested for Cambridge Analytica in Major Data Breach." The Guardian, March 17, 2018. The original investigation that broke the Cambridge Analytica story. Carole Cadwalladr's reporting over multiple years on Cambridge Analytica and its relationships with political campaigns in multiple countries provides the most complete journalistic account of the psychographic targeting scandal. The Guardian and Observer published a series of reports simultaneously with the New York Times; both are worth reading.

3. Confessore, Nicholas, and Danny Hakim. "Data Firm Says 'Secret Sauce' Aided Trump; Many Scoff." New York Times, March 6, 2017. Published a year before the Cambridge Analytica scandal broke fully, this early investigation scrutinized Cambridge Analytica's claims about its effectiveness. Valuable for understanding both the firm's actual capabilities and the role of marketing claims in AI product promotion.

Academic Research

4. Datta, Amit, Michael Carl Tschantz, and Anupam Datta. "Automated Experiments on Ad Privacy Settings: A Tale of Opacity, Choice, and Discrimination." Proceedings on Privacy Enhancing Technologies 1 (2015): 92-112. The foundational empirical study demonstrating that Google's ad delivery system showed higher-paying job ads to men significantly more often than to women, without any advertiser instruction to discriminate by gender. This paper established the research methodology for testing for algorithmic advertising discrimination and is essential for understanding the algorithmic discrimination literature.

5. Ali, Muhammad, Piotr Sapiezynski, Miranda Bogen, Aleksandra Korolova, Alan Mislove, and Aaron Rieke. "Discrimination through Optimization: How Facebook's Ad Delivery Can Lead to Skewed Outcomes." Proceedings of the ACM on Human-Computer Interaction 3, no. CSCW (2019): 1-30. Academic companion to the ProPublica investigations, this paper demonstrates through controlled experiments that Facebook's ad delivery algorithm produces demographically skewed delivery based on ad creative content, independent of advertiser targeting choices. Essential for understanding the mechanism by which algorithmic optimization produces discriminatory outcomes.

6. Mathur, Arunesh, Gunes Acar, Michael J. Friedman, Elena Lucherini, Jonathan Mayer, Marshini Chetty, and Arvind Narayanan. "Dark Patterns at Scale: Findings from a Crawl of 11K Shopping Websites." Proceedings of the ACM on Human-Computer Interaction 3, no. CSCW (2019): 1-32. The most comprehensive empirical study of dark patterns in e-commerce. Documents over 1,800 instances of dark patterns across 11,000 shopping websites, providing the empirical foundation for FTC enforcement attention to dark patterns. Required reading for anyone studying AI marketing manipulation.

7. Kosinski, Michal, David Stillwell, and Thore Graepel. "Private Traits and Attributes Are Predictable from Digital Records of Human Behavior." Proceedings of the National Academy of Sciences 110, no. 15 (2013): 5802-5805. The Cambridge University paper that Cambridge Analytica's approach built on. Demonstrates that Facebook "likes" can predict a range of personal attributes including political views, sexual orientation, personality traits, and more. Essential background for understanding the scientific basis (and limitations) of psychographic targeting.

8. Lambrecht, Anja, and Catherine Tucker. "Algorithmic Bias? An Empirical Study of Apparent Gender-Based Discrimination in the Display of STEM Career Ads." Management Science 65, no. 7 (2019): 2966-2981. An important methodological counterpoint to the discrimination literature: this study finds that gender-based differences in STEM job ad delivery are explained by the higher relative cost of showing ads to women (who receive more advertising overall) rather than by discriminatory intent or bias. A useful reminder that algorithmic disparities require careful causal analysis.

9. Federal Trade Commission. "Bringing Dark Patterns to Light." FTC Report, September 2022. The FTC's comprehensive analysis of dark patterns in digital commerce, providing the regulatory framework for enforcement actions against manipulative design. Covers the range of dark pattern types, their prevalence, their harms, and the FTC's enforcement approach. Accessible and well-organized.

10. Department of Housing and Urban Development, et al. v. Facebook, Inc. Settlement Agreement (2019). The text of the HUD settlement with Facebook. The settlement terms and the accompanying analysis of Facebook's advertising system's discriminatory practices are essential primary sources for understanding how civil rights law applies to algorithmic advertising.

11. Information Commissioner's Office. "Investigation into the Use of Data Analytics in Political Campaigns." ICO Report, November 2018. The UK information commissioner's comprehensive investigation of Cambridge Analytica and the broader use of personal data in political campaigning. The most thorough governmental analysis of the Cambridge Analytica scandal and the data practices it relied on. Includes recommendations for reform that influenced subsequent EU regulation.

12. European Data Protection Board. "Guidelines 02/2022 on the Application of Article 65(1)(a) GDPR." 2022. While technically addressing GDPR consistency mechanisms, this document contains important guidance on the application of GDPR to advertising platforms. Combined with the EDPB's earlier guidance on consent in digital advertising, provides the framework for GDPR-compliant behavioral advertising in the EU.

Books

13. Zuboff, Shoshana. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs, 2019. The most influential academic account of how behavioral data has become the foundation of a new economic logic — surveillance capitalism — in which the prediction and modification of human behavior is the commodity that platforms sell to advertisers. Dense but essential for understanding the structural context of AI advertising ethics.

14. Foer, Franklin. World Without Mind: The Existential Threat of Big Tech. Penguin Press, 2017. A more accessible treatment of the power of technology platforms over commerce, information, and attention. Foer's account of how AI recommendation and advertising systems shape what we see and what we buy provides important context for advertising ethics discussions.

15. Wylie, Christopher. Mindfck: Cambridge Analytica and the Plot to Break America.* Random House, 2019. Written by Cambridge Analytica's whistleblower, this first-person account provides an inside view of the firm's operations, methods, and culture. Wylie's account of how psychographic targeting was developed and deployed is both more granular and more complex than journalistic accounts, and raises important questions about the ethics of working in organizations deploying manipulative AI.

Policy and Advocacy

16. Upturn. "Leveling the Platform: Real Fairness for Advertisers on Facebook." Upturn, 2018. A civil rights technology organization's analysis of the mechanisms of discriminatory advertising on Facebook and the limitations of the platform's proposed remedies. Provides a rigorous civil rights perspective on algorithmic advertising discrimination.

17. Center for Democracy and Technology. "Analysis: EU Regulation on Transparency and Targeting of Political Advertising." CDT, 2022. An analysis of the proposed EU political advertising regulation, its scope, potential effectiveness, and relationship to free speech principles. Useful for understanding the regulatory debate about AI political advertising beyond the US context.

18. Interactive Advertising Bureau (IAB). "Privacy-Preserving Advertising Initiative: Technical Summary." IAB, 2023. The industry's primary effort to develop advertising technologies that reduce reliance on individual behavioral tracking following GDPR, the end of third-party cookies, and growing privacy regulation. Provides the industry perspective on the transition from behavioral to alternative advertising models, including contextual advertising and privacy-preserving measurement.