The Law and Business of Social Media
January 13, 2016 - FTC, Data Security, Privacy, Big Data, Compliance

Big Data, Big Challenges: FTC Report Warns of Potential Discriminatory Effects of Big Data

Big Data, Big Challenges: FTC Report Warns of Potential Discriminatory Effects of Big Data

In a new report, the Federal Trade Commission (FTC) declines to call for new laws but makes clear that it will continue to use its existing tools it to aggressively police unfair, deceptive—or otherwise illegal—uses of big data. Businesses that conduct big data analytics, or that use the results of such analysis, should familiarize themselves with the report to help ensure that their practices do not raise issues.

The report, titled “Big Data: A Tool for Inclusion or Exclusion? Understanding the Issues” grew out of a 2014 FTC workshop that brought together stakeholders to discuss big data’s potential to both create opportunities for consumers and discriminate against them. The Report aims to educate businesses on key laws, and also outlines concrete steps that businesses can take to maximize the benefits of big data while avoiding potentially exclusionary or discriminatory outcomes.

What Is “Big Data”?

The Report explains that “big data” arises from a confluence of factors, including the nearly ubiquitous collection of consumer data from a variety of sources, the plummeting cost of data storage, and powerful new capabilities of drawing connections and making inferences and predictions from collected data. The Report describes the life cycle of big data as involving four phases:

  • Collection: Little bits of data are collected about individual consumers from a variety of sources, such as online shopping, cross-device tracking, online cookies or the Internet of Things (i.e., connected products or services).
  • Compilation and Consolidation: The “little” data is compiled and consolidated into “big” data, often by data brokers who build profiles about individual consumers.
  • Data Mining and Analytics: The “big” data is analyzed to uncover patterns of past consumer behavior or predict future consumer behavior.
  • Use: Once analyzed, big data is used by companies to enhance the development of new products, individualize their marketing, and target potential consumers.

The Report focuses on the final phase of the life cycle: the use of big data. It explores how consumers may be both helped and harmed by companies’ use of big data.

Benefits and Risks of Big Data

The Report emphasizes that, from a policy perspective, big data can provide significant opportunities for social improvements: big data can help target educational, credit, health care, and employment opportunities to low-income and underserved communities.  For instance, the Report notes that big data is already being used to benefit underserved communities, such as by providing access to credit using nontraditional methods to establish creditworthiness, tailoring health care to individual patients’ characteristics, and increasing equal access to employment to hire more diverse workforces.

On the flipside, however, a powerful tool for social good can also be used to discriminate. Just as big data can enhance inclusion, it can also create exclusion. The Report raises concerns that big data analytics may be inadvertently used to exclude certain populations, due to incomplete or inaccurate data, or hidden biases in the collection, analysis, and interpretation of the data. Data may well show correlations that are completely spurious, and if companies base marketing choices on such correlations, unintended harm to consumers may result. The Report provides the example of a credit card company lowering a customer’s credit limit based not on that customer’s payment history, but rather on the fact that the customer shopped at establishments where individuals with poor credit histories had also shopped. In addition, the Report expresses concern that the use of big data may assist in the targeting of vulnerable consumers for fraud, result in higher-priced goods and services for lower-income communities, and exclude such communities from certain offerings.

Maximizing Benefits While Minimizing Risks

Despite recognizing the potential pitfalls of big data, the Report in no way discourages companies from using it. Rather, it seeks to help companies navigate the challenge of how to use big data in a way that maximizes the benefits to them and to society as a whole, while minimizing legal and ethical risks.

  1. Compliance with Potentially Applicable Laws

Companies should understand the laws that may apply to big data practices: specifically, the Fair Credit Reporting Act (FCRA); federal equal opportunity laws, including the Equal Credit Opportunity Act (ECOA) and equal employment opportunity laws (Title VII of the Civil Rights Act of 1964, the Americans with Disabilities Act, the Age Discrimination in Employment Act, the Fair Housing Act, and the Genetic Information Nondiscrimination Ac); and Section 5 of the FTC Act.

  • The FCRA imposes obligations on companies that compile, sell, or use consumer reports. Recent FTC enforcement actions, including United States v. Spokeo, Inc., and United States v. Instant Checkmate, Inc., demonstrate that the FCRA extends beyond traditional credit bureaus to data brokers that compile nontraditional information, including social media information, if that information is used for consumer eligibility decisions. If a company uses big data products to make eligibility decisions (such as to determine a consumer’s eligibility for credit, employment, insurance, or housing), it should make sure that it has complied with all applicable FCRA requirements, including certifying that it has a permissible purpose for obtaining the information and that it will not use the information to violate equal opportunity laws. Similarly, the company must provide consumers with any notices required under the FCRA (such as notice of adverse action taken based on information in a consumer report) and obtain any required authorizations (such as to procure a consumer report for employment purposes).
  • Federal equal opportunity laws prohibit discrimination based on protected characteristics such as race, color, sex, religion, age, disability, national origin, marital status, and genetic information. Discrimination may take the form of disparate treatment (intentionally treating an individual differently based on a protected characteristic) or disparate impact (applying a facially neutral policy that has a disproportionate adverse effect on individuals with a protected characteristic). The Report suggests, for example, that it may be problematic under the ECOA for a company to make credit decisions based on consumers’ zip codes, if the decisions have a disproportionate adverse impact on a particular ethnic group and are not justified by a legitimate business necessity. Companies should thus examine whether their use of big data results in the treatment of people differently on a prohibited basis, whether directly or indirectly, and take steps to remedy any such discrepancies.
  • Section 5 of the FTC Act prohibits unfair or deceptive acts or practices. Companies engaging in big data analytics should consider whether they are violating any material promises to consumers, such as a promise to refrain from sharing consumer data or an assurance that they will safeguard it. Further, companies that supply big data to other companies should undertake reasonable measures to know the purposes for which their customers will use the data. For example, companies should take steps to ensure that their customers will not use big data to commit fraud or for discriminatory purposes.

The Report urges companies to proceed with caution when using big data in ways that could lead them to violate these and other potentially applicable laws.

  1. Policy Considerations for Big Data Research

The Report urges companies to take into account the special policy considerations raised by big data research. To help minimize the potential for discriminatory harm arising out of such research, the Report encourages companies to address the following questions:

  • How representative are your data sets? Consider whether they are missing information from particular populations and, if so, take steps to address issues of under- and overrepresentation.
  • Does your data model account for biases? Review your data sets and algorithms to ensure that hidden biases do not have an unintended impact on certain populations.
  • How accurate are your predictions based on big data? Remember that just because your big data analytics found a correlation, the correlation is not necessarily meaningful. Human oversight of big data tools may be worthwhile, particularly when decisions implicating health care, credit, or employment are involved.
  • Does your reliance on big data raise ethical or fairness concerns? Consider whether such concerns advise against using big data in certain circumstances. Consider further whether you can use big data to advance opportunities for underserved populations.


Particularly in light of ever-developing technologies, big data will certainly continue to grow in importance. The Report makes clear that the FTC will continue to monitor companies’ legal obligations around its use. Accordingly, companies should carefully examine their big data practices to identify and take steps to help minimize the risks they may present.