Civil Rights Enforcement in the Era of Big Data: Algorithmic Discrimination and the Computer Fraud and Abuse Act

Welcome to the “Big Data” Revolution. Educators use data to improve learning outcomes. Marketers use it to predict fashion trends, forecast demand, and optimize pricing. And police use it to map crime before it happens. Indeed, the World Economic Forum even declared data to be “a new asset class,” akin to oil. Despite the meteoric rise of Big Data over the past two years, the quantity of data generated to date may pale in comparison to the amount that will be generated once companies add sensors to their products. Although the explosion of personal data brings with it the possibility of transformative change, it also houses the potential to impact classes of people in discriminatory ways. Increasingly, transactions involving the core social goods of civil rights laws – housing, employment, credit, and consumer transactions – take place online. Meanwhile, discrimination persists in these markets. In an era of Big Data that brings tremendous possibilities and frightening perils, it is time for Congress to reconsider the future enforcement of antidiscrimination laws.This Note will be the first to address the impact of the federal government’s principal anti-hacking statute – the Computer Fraud and Abuse Act (CFAA) – on enforcement of civil rights in online markets. It will begin by looking at American society’s production of personal data, and American companies’ reliance on proprietary algorithms to “mine” this data in order to make business decisions. It will then detail examples of algorithmic steering and targeting, and their potential to violate civil rights laws in online markets. In particular, online real estate databases – like Zillow.com and Realtor.com – represent one such marketplace where these violations may occur. When a consumer searches for available housing, the website’s algorithms may not provide the same housing options, or may steer the user toward different options, based upon a protected classification like race. While algorithms can be programmed to discriminate based on protected classes, algorithms may also unintentionally use personal data – such as the consumer’s ZIP code and online purchase history – as a proxy for a protected class. Both produce actionable discrimination. Thus, seemingly neutral algorithms can discriminate in ways analogous to those of real world real estate agents. While this problem will be analyzed in the context of the Fair Housing Act (FHA), the analysis is applicable to other civil rights laws that are enforced through audit-based testing – for example, the Equal Credit Opportunity Act and Title VII of the Civil Rights Act of 1964. Part I of this Note will provide background on the rise of Big Data and demonstrate how American corporations use data to inform business decisions. It will detail some ethically questionable uses of data, and demonstrate how such usage has the potential to discriminate unlawfully against protected classes of people. Part II will demonstrate the central role of audit-based testing in the enforcement of civil rights laws, specifically the FHA. It will then outline best practices for conducting analogous audits of online real estate websites, and discuss a legal barrier to such audits – the CFAA. In order to combat discrimination in online markets, one must look to the development of legal protections against discrimination in offline markets. Part III will offer examples of past reform efforts to the CFAA, detail the shortcomings of these efforts for the purpose of auditing algorithms, and propose potential remedies available to federal courts and Congress. While this Note will be the first to address the effect of the CFAA on online enforcement of civil rights laws, it will also contribute to the ongoing discussion of protecting civil rights in the Information Age.