How to Mitigate AI’s Exacerbation of Income Inequality

Haran Segram
Quote icon
To keep the American Dream alive, policymakers must address AI’s exacerbation of income inequality.
By Haran Segram
It is forecasted that approximately one-sixth of the total labor force will be made redundant over the next decade through some form of automation; this number will be likely be higher in developed economies, including the United States. To keep the American Dream alive, policymakers must address AI’s exacerbation of income inequality.

For workers with jobs that are repetitive in nature, the transition into the AI age will be a particularly painful one. For example, Amazon Go plans to open 3000 “cashierless” stores in the next three years. Additionally, the advent of digital wallets will make several categories of traditional banking jobs redundant. Transportation via autonomous vehicles, although it will lower transportation costs and possibly improve safety, will lead to disruption and dislocation of insurance and transportation workers. Eventually, medical technicians could also potentially be replaced by robots. Even with substantial gains in new categories of employment, wealth distribution could veer beyond an 80/20 model, moving us closer towards a “winner take all” society.

The good news is that AI will create new jobs with titles unheard of previously, but they will require workers with specialized skills and training. Through education, early adopters will reap the maximum benefits of AI. While projects like AI4ALL are encouraging beginnings, more widespread action is needed. Government investments, subsidies and private-public partnerships are paramount in enabling citizens to adapt to this changing employment landscape. The Presidential executive order on AI is the first significant move to combat AI-induced inequality by offering training opportunities through the Select Committee on Artificial Intelligence and the National Council for the American Worker. Funding technological education and forgiveness of student debt by closing the tax loopholes, such as the carried-interest loophole, and broadening the tax base could mobilize underprivileged individuals to climb the social ladder.

Additionally, courses on AI, Machine Learning, Deep Learning and Blockchain technologies should be introduced to high school students. Vocational study centers should be made available and affordable. Training in these subject areas, led by qualified teachers, should be as commonplace as apprenticeships for electricians and plumbers. A mix of online learning and physical classroom teaching might be one possible solution.

AI presents complex legal challenges that require policies that simultaneously protect data privacy without hindering technological innovation. Regulators and policymakers should be both proactive and sensitive to the way we live and work in the AI age. While overreaching regulation will curb the growth and further development of AI, regulations and policies should be developed for its unintended consequences. For example, who is responsible in the case of a car accident: the driverless car manufacturer or the owner of the car? If a serious error occurs during a surgery, is the surgeon or the maker of the co-working intelligent agent responsible, or both? Lawmakers, such as those who questioned Mark Zuckerberg regarding privacy and data misuse, need an in-depth command of data privacy issues. To deter future scandals like Cambridge Analytica and to prevent “friendly frauds,” governments need to hire more technologists, such as AI and cybersecurity specialists, at competitive salaries. This will increase the cost for government in the short term but will offer distinct long-term benefits.

In a recent simulation study, McKinsey estimated that AI could account for 1.2 percent additional growth annually to the global GDP over the next decade. In other words, $13 trillion will be added to the global GDP -- approximately twice the ‘internet era’ value add of 0.6 percent per year. However, GDP does not necessarily reflect the wellness of a population. Without innovative policy measures, income inequality will be further exacerbated, potentially harming our most vulnerable populations. In the United States, we should not rely on the invisible hand to lessen AI-driven income inequality.

Haran Segram is a Clinical Assistant Professor of Finance. 

Professor Segram wishes to sincerely thank Kimberly Couzens at NYU Stern Public Affairs for her excellent editorial support.

Big Ideas 2019’, Ark Invest 2019

Brynjolfsson, E. and McAfee, A., ‘The Second Machine Age: Work, Progress, and Prosperity in a time of Brilliant Technologies’, W.W.Norton & Company Inc. 2014

Ford, M., ‘Architects of Intelligence: The truth about AI from the people building it’, Packt Publishing 2018

Pilling, D., ‘Growth Delusion: Wealth, Poverty and the Well-Being of Nations’, Penguin Random House 2018;

Wilkinson, R. and Pickett, K., ‘The Inner Level: How more equal societies reduce stress, restore sanity and improve everyone’s well-being’, Penguin Press 2019