Initiative NO. 1


Algorithmic Safety: Mitigating Bias in Workforce Decisions

200 +
experts from business, academia,
and civil society helped test and refine the safeguards.

Businesses and institutions are increasingly applying data, algorithms, and AI to support their workforce decisions—from hiring and promotion to productivity and compensation.

These technologies help companies identify talent in larger and more diverse pools of candidates, better match the right talent to the right opportunity, personalize employee experiences, and automate routine tasks to free up time for more meaningful work.

However, the technology’s ability to support the workforce also comes with the risk of potentially worsening unfair bias. Alliance member organizations identify unfair bias as one of the highest risks when using these technologies to augment workforce management.

Most of the algorithmic systems used to support workforce decisions are introduced and maintained by vendors—including software providers, professional networking sites, benefits planners, and recruiting firms.

Therefore, we have developed Algorithmic Bias Safeguards for Workforce—criteria and education for HR teams to evaluate vendors on their ability to detect, mitigate, and monitor algorithmic bias in workforce decisions.

the Overview

The safeguards

How were the Safeguards created?

The Safeguards were developed by a working group of member company experts in human resources; artificial intelligence; IT; procurement; legal; AI model evaluation; and diversity, equity, and inclusion. They were then tested and refined with input from vendors, academia, civil society organizations, and subject matter experts in business. In total, more than 200 experts were involved in the process.

What do the Safeguards include?

The Algorithmic Bias Safeguards for Workforce include four components:


55 questions in 13 categories for completion by the HR vendor. Answers are not shared among member companies.


Education and Assessment
To enhance algorithmic literacy and to provide detailed guidance for HR teams assessing vendor responses to the Evaluation.


To qualitatively grade and compare vendors and document issues.


Implementation Guidance
For integrating the Safeguards into an organization’s systems. The Safeguards supplement member companies’ vendor selection procedures.
This is not only an anti-bias tool. It is an improved outcomes tool for a time in which talent recruitment and retention are becoming critical for fairness, opportunity, and future business success.”
— Nuala O’Connor, SVP, Chief Counsel, Digital Citizenship, Walmart

As of December 8, 2021, 13 members of the Alliance have committed to adopting version 1.0 of the Algorithmic Bias Safeguards for Workforce.

As organizations implement version 1.0 of the Algorithmic Bias Safeguards for Workforce, we will learn more through real-world use—with the goal of evolving the Safeguards so they can be shared widely.

Download an overview of the Safeguards here.

Please contact us if you would like to review the Algorithmic Bias Safeguards for Workforce version 1.0 and learn more.

the Overview