Despite current legal challenges to diversity, equity, and inclusion policies and the rise of artificial intelligence, AI won’t spell the end of DEI initiatives. Sure, we might end up with a watered-down version, but DEI will survive.

To understand how AI impacts DEI in business, we need to consider fundamental stereotypes—the ones that stubbornly stick around. For example, as a female attorney, here are some stereotypes that I confront. Do women practice law differently than men? Swap out law for any field – medicine, finance, engineering – and the question remains: do women perform these jobs differently? Are female professionals more aggressive to counteract stereotypes? More efficient because they multitask? More compassionate (this stereotype makes me cringe, but it’s persistent)?

These questions highlight biases—some unconscious, some blatant. Depending on the recruiter’s viewpoint, teams may end up female-dominated, sidelining male and LGBTQIA+ applicants, or underrepresenting women altogether.

Enter artificial intelligence. AI mimics human thinking to analyze data and make decisions. Pre-AI, we had two simple controls to address hiring biases: visual observation and lawsuits. HR or managers would observe trends and potentially address them, and if not, an applicant might sue.

But with AI, things get tricky. Visual observation is tough in large companies, and employers can’t see how AI processes data. Consequently, AI might exclude diverse applicants without anyone realizing it, depriving organizations of fresh perspectives.

Indeed, for example, uses automation and algorithms to match job seekers with employers, but doesn’t make employment decisions. However, AI’s role in the hiring process still raises questions. Does AI introduce unconscious bias in employment decisions? Meta AI says yes—AI can have unconscious bias because it’s trained on data that may be biased. This can lead to discrimination, inaccurate results and the perpetuation of stereotypes.

AI is now a key player in hiring, succession planning and performance management. Will AI’s unconscious bias create a workforce that undermines diversity? It’s possible—AI learns human biases along with human thinking.

We can control bias through litigation, regulation and corporate policies. Companies should continue to demand diversity from their business partners. Laws are emerging at local and state levels to regulate AI in employment, and the EEOC has issued guidance for employers on using AI. Discrimination lawsuits related to AI in hiring have already started, like Mobley v. Workday Inc.

Consider how AI is used in different industries. In finance, AI algorithms are employed to predict market trends and manage risk. However, if these algorithms are based on biased data, they may inadvertently favor certain groups over others. In health care, AI helps in diagnosing diseases and recommending treatments. But biased training data can lead to misdiagnoses and unequal treatment recommendations.

A striking example is Amazon’s AI recruiting tool, which was scrapped after it was found to be biased against women. The tool, trained on resumes submitted over a decade, favored male candidates because most of the resumes came from men, reflecting the industry’s gender imbalance. This incident underscores the importance of scrutinizing AI systems for bias and ensuring they are trained on diverse and representative data sets.

Another area of concern is facial recognition technology, which has been shown to have higher error rates for people with darker skin tones. When such technology is used in hiring processes or employee monitoring, it can result in unfair treatment and discrimination.

To counteract these issues, organizations must adopt a multifaceted approach. First, they should invest in bias detection and mitigation tools to regularly audit their AI systems. Second, transparency is crucial; companies must be open about how their AI systems work and the data they use. Third, involving diverse teams in the development and deployment of AI can help identify and address potential biases.

The role of human oversight cannot be overstated. AI should assist, not replace, human decision-makers. Human oversight ensures that decisions are contextually appropriate and ethically sound. This hybrid approach leverages the strengths of both AI and human judgment, minimizing the risk of bias.

Education and training are also vital. Employers must educate their workforce about AI and its implications for DEI. Training programs should focus on recognizing and addressing biases, both human and algorithmic.

Finally, collaboration between the public and private sectors is essential to develop standards and regulations that promote fair and unbiased AI. Initiatives like the Partnership on AI, which brings together companies, researchers and civil society organizations, are steps in the right direction.

While AI poses challenges to DEI, it also offers opportunities for improvement. By proactively addressing bias, promoting transparency and fostering collaboration, we can harness AI’s potential to enhance rather than hinder diversity, equity, and inclusion. The future of DEI depends not on the demise of AI but on our ability to integrate it responsibly and ethically into our systems and practices. The big question is: will these measures be enough to keep DEI meaningful? Time will tell, but one thing’s for sure—AI isn’t going to kill DEI.

Brigette Eagan is a partner at Newark’s Genova Burns and leads the firm’s Human Resource and Compliance Group.

Editor’s note: This opinion column was updated at 10:27 a.m. July 30, 2024, to correct the spelling of Brigette Eagan’s name.