Comments to the EEOC on the Implications of Big Data in Employment
The Center for Democracy & Technology (CDT) respectfully submits these comments in response to the Commission’s October 13 public meeting on the implications of big data in employment. CDT is a nonprofit public interest organization dedicated to promoting digital privacy, free expression, and individual liberty. CDT commends the Commission’s effort to protect equal opportunity in employment by examining how the collection and use of data may exacerbate structural inequality. In particular, we urge the Commission to scrutinize the use of automated decision-making systems in hiring, management, and employee evaluation practices and to ensure that such systems promote fairness, equality, and diversity.
Can algorithms deliver on their promises to hire more successful employees, improve office culture, and increase productivity? Responsible algorithms could be a force for good in the modern workplace, but the risks of harm that attend automated hiring, managing, and assessment systems are manifold. Rather than offering unbiased alternatives to human subjectivity, algorithms are imbued with the values of those who create them. The natural inclination to hire people who look and act like us can be encoded and perpetuated by algorithms that seek to recreate historical patterns. This bias toward past hires is likely to disproportionately favor historically advantaged groups — such as affluent, white, college educated males — to the detriment of historically disadvantaged groups. Moreover, the trend toward collecting and analyzing large amounts of data on existing employees — whether to optimize performance or improve morale — threatens to erode important barriers between personal life and work life. These comments recommend steps the EEOC can take, and principles it can follow, in order to mitigate the negative consequences and support the positive impacts of big data and automated decision-making in the workplace.
As concern about risks and harms related to AI systems continue to grow, a growing chorus of policymakers, industry leaders, and advocates have called for independent AI assessments. This explainer provides an overview of recent proposals for third-party assessment in the United States, including state and federal legislation, executive actions, and industry proposals.
Having third parties assess AI systems might seem like common sense, but crafting effective policies toward this goal can be devilishly tricky. A poorly-constructed ecosystem for third-party assessment could easily fail to consider the most consequential mechanisms of risk, neglect the AI harms that most impact people, or do more to protect AI companies than people.
Not All Guardrails Are Created Equal: Comparing Content Safety and Copyright Filtering
As courts and policymakers work through questions about chatbot liability, they should be wary of analogies that flatten meaningful technical differences. Copyright filtering and safety intervention share real challenges around ambiguity and evasion, but they diverge in what each control must assess, how each manifests over the course of a conversation, and how much can be verified from the outside.
