On Nov. 2, 2018, we discussed a lawsuit against Facebook alleging discriminatory big data practices. In March 2019, just six months after the ACLU filed charges with the EEOC, Facebook settled the Title VII claims against it for approximately $5 million.

As part of the settlement, in addition to this significant cash payout, Facebook agreed to overhaul its entire targeted advertising system to withhold detailed demographic information. Specifically, Facebook will withhold information related to gender, age and zip codes from certain advertisers, including those advertising employment opportunities. Facebook's overhaul is intended to hinder advertisers' ability to micro-target specific demographics, which, in turn, should help minimize discriminatory advertising practices the company is accused of perpetrating.

Although Facebook's settlement means a court will not decide the merit of the ACLU's claims, understanding Facebook's alleged misconduct provides a playbook for any business looking to capitalize on big data's various applications, while avoiding its inherent risks.

Capitalizing on Facebook's Alleged Missteps

While Facebook denied the allegations, the settlement terms are significant enough to warrant the attention of any company looking to utilize big data algorithms.

Businesses who fail to harness the power of big data risk falling behind competitors. However, those looking to use big data applications to assist in talent recruitment and other employment practices can learn from the mistakes of others already employing big data and avoid issues presented by big data algorithms. Whether a company plans to use big data analytics for advertising, recruiting, employee performance monitoring, business operating costs, or the many other potential big data applications, having a strategic big data plan can help mitigate risk on the front end and reap valuable returns on the back end.

The EEOC's Pay Data Requirements

For example, a company could use big data analytics to compile employee wage and pay data. In fact, employers with more than 100 employees are required annually to submit wage and hour data organized by job category, gender, race, and ethnicity, classified as EEO-1 Component 2 data to the EEOC. EEO-1 Component 2 is a submission to the EEOC that includes year-end earnings and hours worked grouped by EEO-1 job category and separated by race/ethnicity, gender, and pay band.

By collecting this data, the EEOC hopes to ensure employees are receiving equal pay for equal work. In future discrimination allegations, having access to wage and hour data allows the EEOC to better assess the company and the situation. However, companies should be mindful that submission of this material to the EEOC could be scrutinized, particularly if wage and hour data evidences pay disparities among employees in different protected classes.

On Sept. 11, 2019, the EEOC announced that it would not seek collection of calendar year 2019 EEO-1 Component 2 data. See H. Mark Adams, EEOC Cans Component 2 Pay Data Collection Rule … After September 30, The National Law Review (Sept. 13, 2019). Despite this new development, the EEOC just finished collecting EEO-1 Component 2 data on September 30, 2019. As a result, employers should prepare themselves as the EEOC sorts through employer's recent submissions.

Big Data in Hiring Practices

Companies may use big data applications to find the ideal job candidates, as these algorithms can identify those applicants who gave a better chance of a positive job performance.

However, taking advantage of big data to assist when seeking top talent can also pose serious risks for companies. While algorithms can help employers target their recruiting efforts to find the ideal candidate for a specific position, if an employer does not scrutinize their hiring processes, they could find themselves facing employment discrimination claims.

Proactively Analyzing Data Mitigates Risk

By proactively compiling and analyzing this information via big data analytics before a lawsuit or submission to the EEOC, a company can spot potential red flags indicating disparate wage practices based any number of protected classes including race, color, religion, sex or national origin. Performing a "self-audit" and identifying these red flags can help a company factor in neutral, job-related employee performance metrics, like job grades; service time; and employees' talent scores for past performance and future potential. In turn, that company would be in a better position to defend itself against future discrimination claims and avoid facing the same fate as Facebook if the EEOC or a jury determines that the company's 2017 or 2018 EEO-1 Component 2 data evidences pay disparities among employees in different protected classes.

Put another way, the company who strategizes exactly how it will implement big data analytics can take proactive measures to establish safeguards to effectively shield itself from future discrimination claims. Early involvement between counsel and technology specialists can help optimize risk avoidance.

In conclusion, although Facebook's settlement agreement stamps out its potential discrimination exposure for the time being, companies should be on the lookout: Additional lawsuits are inevitable as companies push the boundaries lawmakers establish around this rapidly expanding legal and technological paradigm. Application of existing laws to big data analytic use, and the lawsuits that result therefrom, can only clarify acceptable big data usage for future big data utilizers.

Bret Cohen is a partner with Nelson Mullins and chairs the labor and employment practice. Robert O. Sheridan is a partner and Timothy M. Harvey is an associate at the firm. They can be reached at [email protected], [email protected] and [email protected], respectively.