What will generative AI mean for the racial wealth gap? 

Big technological shifts have a history of keeping Black and other marginalized groups at or near the bottom of economic opportunities. Will this time be different?

SOURCECenter for Public Integrity

Kelcey Gibbons, a doctoral student who studies African Americans in technology and the Black middle class, is not quite sure what she makes of generative artificial intelligence and how it might impact the racial wealth gap. 

Gibbons anticipates that generative AI will force organizations to rethink which skills matter in the workplace, aggravating existing inequalities without a strong emphasis on access and education. 

It could also become a scapegoat for why disadvantaged groups struggle with upward mobility. Whether it narrows the racial wealth gap will depend on participation from the bottom up, Gibbons said.

Generative AI is a broad term for artificial intelligence that can create new content, such as art, audio, code, music, text and videos. It uses machine learning, another form of artificial intelligence that gleans patterns from data and algorithms that help to process new and more information.

The technology can increase productivity by automating tasks now done by humans. For example, chatGPT is a tool trained to produce text and code. It can provide detailed responses to questions, translate languages, identify errors in code and provide copywriting in the voice of a brand.  

Other generative AI tools and platforms can create interactive virtual learning or gaming environments. Professional sports leagues like the NBA and NHL are using generative AI to provide more detailed sports content to their consumers and better understand their audiences.

Workers fear generative AI will replace human jobs, which will affect Black people, who have the highest unemployment rates in the nation. 

In other words, generative AI is just the latest version of the old saying: “When America catches a cold, Black people get pneumonia.” In this case, it’s because Black people are disproportionately in the jobs that technology is likely to replace. And it could mean, once again, that fewer Black people will have decent jobs. 

According to a recent report by the McKinsey Institute for Black Economic Mobility, part of the business management and consulting firm McKinsey & Co., generative AI has the potential to widen the racial wealth gap between Black and white Americans by $43 billion annually over the next 20 years. 

But if generative AI is applied with equity in mind, it could instead help close the racial wealth gap and remove barriers to Black economic mobility, the report says. 

Not so fast, said Amy Slaton, a professor of history at Drexel University in Philadelphia. Despite discourse around inclusion and a “tremendous faith in the idea of merit,” the systems in which people of color and marginalized groups operate seldom change.  

“Generative AI comes in with a high-tech promise for the economy, but it’s not,” Slaton said. “Nothing is changing for the communities that make up that economy.”

Big technological shifts have a history of keeping Black people at or near the bottom already.

During the 17th, 18th and 19th centuries, global trade networks yielded tremendous profits from buying and selling human beings. As global trade, including cotton and sugar production, grew, so did regimes of enslavement.

Slavery ended. Agriculture became more mechanized. New industrial opportunities popped up in cities amid a massive scale-up in manufacturing and consumption.

Black people exercised agency in their lives by migrating from the rural South to economic opportunities in urban areas in the North. But as companies developed assembly lines, new managerial techniques and later automation, reducing jobs, they did so in service to white capitalists, Slaton said. 

“Capitalism in America is racial capitalism,” Slaton said.

To be sure, the Civil Rights era and integration ushered in sweeping changes for disadvantaged groups. “Individuals’ lives are drastically changed,” Slaton said. “But it’s a hierarchical economy, and people of color are historically at the bottom. AI will exacerbate that.” 

According to the McKinsey report, Black workers are overrepresented in four of the top five occupations most at risk of automation. They include office support, production work, food services and mechanical installation and repair. Roughly three in four Black workers lack a college degree. 

But in the past five years, one in every eight Black workers without a four-year degree has moved to a job that rewards experience, offers annual salaries above $42,000 and provides a path of upward mobility. These jobs include training and development specialists, and sales managers. 

The problem is that these paths are likely to disappear, too. Between 2030 and 2060, the report says, generative AI is expected to replace about half these jobs that don’t require college degrees. 

To avoid being displaced, the report argues that workers should gain skills that machines cannot do. Those “future-proof skills” include “socioemotional understanding” and “comfort with ambiguity.” Occupational health and safety specialists and special education teachers are two examples of occupations that use those skills.

In the tech industry, workers should not merely know how to use a “specific coding language to code formulaically,” but should instead understand “computational and statistical principles in order to troubleshoot and solve problems with machine-generated code,” the report says.

“Capitalism in America is racial capitalism.”


That’s true for any technological development, said James Pristas, professor in the department of computing and information technology at the College of Southern Nevada in North Las Vegas. Pristas teaches an introduction to artificial intelligence course at the school, a public community college where more than half the student body is Latino, Asian or Black. 

It’s too soon to know how generative AI will shape the workforce, he said. The immediate impact, though, is how students are using it to complete their assignments, which makes it difficult for professors like him to assess them. 

“It’s a major step change that is going to change the way we teach and the way kids learn,” Pristas said.

So can generative AI be deployed in a way that narrows the racial wealth gap? 

“I think that’s a question of power and access,” said Gibbons, a PhD student at the Massachusetts Institute of Technology. “Who gets to say what generative AI is, and for whom?”

Generative AI tools are already used to evaluate job candidates through keyword requirements, multiple choice exams and even voice or video analysis. They risk engaging in employment discrimination by excluding people based on race, religion, gender identity and other protected classes, according to the ACLU.

Last fall, President Biden signed an executive order on developing AI safely and responsibly. And in January, the administration said that major AI companies must now report their safety tests to the government. “My Administration cannot — and will not — tolerate the use of AI to disadvantage those who are already too often denied equal opportunity and justice,” Biden’s executive order says.  

Congress has yet to enact legislation on generative AI. Meanwhile, state and local governments are writing their own rules.

According to McKinsey, generative AI can be used with an equity lens by establishing regulatory guardrails, training it on unbiased and “authentically representative” datasets, growing “diverse tech talent,” engaging all stakeholders in the design of new products and encouraging leaders to use the technology only where it can make unbiased decisions. 

The goal is to “have all people and communities — including Black communities, previously left behind by seismic shifts in tech — benefit from this amazing technology transformation,” the report says in closing. 

“The time to act is now, using the foresight of past transformations as a guide.”


If you liked this article, please donate $5 to keep NationofChange online through November.