������Ƶ

Artificial Intelligence (AI) can help business owners and managers do a growing variety of jobs, from to .

This digitized workplace needs to protect workers. Such guidance can also help software designers, business owners, and managers create AI applications that maximize business benefits and minimize risk for themselves and their workers–especially when it comes to preventing discrimination enabled by AI. Workers need to know their rights if they think an employer has discriminated against them.

This sort of guidance can come from the , the federal government agency established by the Civil Rights Act of 1964 that enforces federal laws against workplace discrimination. To understand how the EEOC is adapting to AI, it is first helpful to know what the EEOC is and what it does.

What Does the EEOC Do?

The EEOC is responsible for enforcing various laws that prevent workplace discrimination, including the , the , and the .

Under all these laws, the EEOC takes two main approaches: first, it helps individual workers and worker organizations, such as labor unions and , understand their rights and helps employers know their duties; second, it investigates and resolves alleged incidents of discrimination.

The EEOC has the authority to investigate and prosecute cases against most organizations, including labor unions and employment agencies, employing 15 workers or more.

EEOC role in Advising Employers

Rulemaking: The EEOC creates rules and legally binding regulations to help it do its job. These rules are examples of because Congress gave the EEOC leeway to make the rules it needs to administer anti-discrimination laws. Making these rules is and includes a public comment period where businesses, nonprofits, and individuals can before a new proposed role takes effect. EEOC regulations can also face challenges in court, including from .

Guidance: Short of making a new rule, the EEOC can publicly announce already in effect. These announcements offer businesses advice on how to comply with the law and what will happen if they do not.

EEOC role in Representing Workers

In 2023, the EEOC secured for private-sector workers who faced discrimination. Behind that number is a multi-step process based on collecting, investigating, and resolving workers’ charges of discrimination before filing lawsuits.

Investigations: When an employee , known as a “charge of investigation,” with the EEOC alleging their employer discriminated against them, the EEOC investigates and decides what to do. Sometimes, the agency suggests , a confidential, non-binding meeting between the worker and employer to resolve disagreements. In 2023, the EEOC filed after 233,704 in-person requests, 522,000 phone calls, and over 86,000 emails from individual workers.

Settlement: If the EEOC determines the employer broke the law, the EEOC will try to the case before the EEOC or the worker files a lawsuit. When the sides agree to settle, a court can enforce their agreement. In recent years, the has been about $40,000.

Litigation: If the EEOC cannot settle a case, they will either sue the employer or give the complaining worker a “” letter that allows them to sue their employer in federal district court. Workers can also request these letters earlier in the process under some laws. The commissioner of the EEOC can also issue charges without a complainant, referred to as a "." In the , the EEOC reported resolving 98 lawsuits in federal district court, winning 91% of them, and recovering a combined $22.6 million for 968 workers.

Does the EEOC have the capacity it needs?

It takes time to follow the process for every case under every law. The EEOC has battled a and limited resources for years. For that reason, the EEOC tries to “on charges where the government can have the greatest impact on workplace discrimination” and on preventing discrimination in the first place through guidance.

How can the EEOC impact workplace AI?

The EEOC believes AI-based discrimination is important. The EEOC has been developing an “,” so more action is surely on the way. A large part of the EEOC’s work is guidance and education, talking to workers and employers about how to navigate AI and producing guidance documents exploring how to prevent AI-based forms of discrimination, such as .

One of the EEOC’s Commissioners, , once wrote that AI in the workplace had both “,” depending on how regulators and businesses succeeded in applying America’s existing laws. Although Commissioner Sonderling wrote that article in a private capacity, the EEOC’s work reflects this same mission. Nick Truxal, Commissioner Sonderling’s Chief Advisor on Human Resources and Artificial Intelligence, told me in an interview that, as public interest in AI has soared, business leaders’ duties remain the same.

For example, if a company wants to use a startup’s AI-based resume-reviewing tool, the hiring company and the startup itself will have to make sure the AI tool does not unintentionally discriminate against job applicants based on their race, gender or other protected parts of their identity. Amazon faced this problem in 2018: its resume-reviewing AI was trained on reading ten years of Amazon’s previous tech-sector hires, which were dominated by men; based on this information, the AI was actively penalizing resumes that used the word “.” The modern EEOC wants to help companies prevent this kind of discrimination by ensuring AI applications are, as Commissioner Sonderling has privately stressed elsewhere, “.”

But this general answer leads to more questions: Will this guidance be enough? How can the EEOC prevent discrimination without stifling the experimentation necessary for innovation? How can the EEOC account for employment discrimination laws in all 50 states and countries all around the world?

������Ƶ will dig deeper into the EEOC’s AI actions during an upcoming event featuring Commissioner Sonderling hosted by ������Ƶ's Future of Work and Innovation Economy initiative.