AI Regulation Beyond AI Laws: HR & Discrimination

Including: 9 laws and 2 regulatory guidances for AI in HR

Hi AI ethics enthusiasts,

When we think about AI regulation, our minds often jump to AI-specific laws like the EU AI Act. This perspective overlooks the fact that AI companies are subject to the entire legal system, including many general laws that apply to AI.

This oversight is dangerous. For individuals, focusing on AI-specific regulation can create an illusion that AI-enabled products and companies are somehow exceptions to the legal system, allowing them to get away with harming people and breaking the law. For companies, focusing on AI-specific laws is too narrow. It creates exposure to compliance and reputational risks.

Therefore, understanding how non-AI-specific laws apply to AI is crucial. To that end, today I will review non-AI-specific laws that apply to the Human Resources (HR) sector in the US. I will focus on anti-discrimination, given how prevalent bias is in AI tools. I also made a cheat sheet for our paid subscribers!

My review is based on Lena Kempe's wonderful article, "Navigating the AI Employment Bias Maze: Legal Compliance Guidelines and Strategies", which I highly recommend reading.

For dessert, an AI-generated take on this post!

Three Categories

The Human Resources (HR) sector is already the target of several AI-specific laws, most notably New York City’s 144, which requires conducting bias audits to automated decision tools for employment. However, the AI-specific laws are only the tip of the iceberg. AI tools are also subject to general laws, including:

• Sector-specific laws

• General laws, such as non-discrimination laws

• Guidance from regulators, such as federal agencies

Moreover, we have already seen lawsuits based on these laws, which I will get to.

HR-Specific Laws

The following fairness-related laws regulate the HR sector in the US. They don’t target AI-enabled companies, but they apply to these companies just the same, and irresponsible use of AI may violate them.

1. Title VII of the Civil Rights Act of 1964 (“Title VII”)

   Prohibits employment discrimination based on race, color, religion, sex 

2. The Equal Pay Act

   Prohibits sex-based wage discrimination.

3. The Age Discrimination in Employment Act

   Prohibits discrimination based on age (forty and over).

4. Title I and Title V of the Americans with Disabilities Act (“ADA”)

   Prohibits employment discrimination against qualified individuals based on disability and those regarded as having a disability.

5. The Pregnant Workers Fairness Act

   Prohibits discrimination against job applicants or employees because of their need for a pregnancy-related accommodation.

6. The Genetic Information Nondiscrimination Act

   Prohibits discrimination in employment and health insurance based on genetic information.

General laws: Anti-discrimination

The HR sector is bound by general laws that AI adoption may compromise, including the following anti-discrimination laws:

1. Section 1981 of the Civil Rights Act of 1866

   Prohibits discrimination based on race, color, and ethnicity.

2. The Immigration Reform and Control Act

   Prohibits discrimination based on citizenship and national origin.

3. The Uniformed Services Employment and Reemployment Rights Act

   Prohibits discrimination against past and current members of the uniformed services, as well as applicants to the uniformed services.

Regulatory guidance: EEOC initiatives

The U.S. Equal Employment Opportunity Commission (EEOC) has recognized the need to provide guidance on how these existing laws apply to AI. They've issued two significant pieces of guidance:

1. EEOC Guidance on AI and the Americans with Disabilities Act (May 2022): This document outlines how AI tools could potentially violate the ADA and suggests practices for responsible AI use to mitigate risks.

2. EEOC Guidance on Title VII of the Civil Rights Act (May 2023): This technical guidance explains how to measure "adverse impact" when AI tools are used in employment selection processes, defining when outcomes may violate Title VII.

Lawsuits

We're already seeing lawsuits based on non-AI-specific laws. A prime example is the Mobley vs Workday case, in which Derek Mobley sued Workday for hiring discrimination, alleging he was turned down from more than 100 jobs that used Workdays' AI platform due to his race, age, and disability status. The case is currently still ongoing.

Cheat Sheet

For ease of reference, here’s a cheat cheat summarizing the laws I reviewed today: