According the Federal Reserve Bank of St. Louis, as of the second quarter of 2024, the top 10% of households controlled 67% of total wealth, averaging $6.9 million per household. In stark contrast, the bottom 50% held only 2.5% of the nation’s wealth, with an average of $51,000 per household.
Joseph Stiglitz is renowned economist and recipient of the Nobel Prize in Economics in 2001. Stiglitz posits the view that AI can worsen existing inequalities and consolidated power in the hands of a few dominant corporations, ultimately undermining economic fairness and democratic stability.
Adam Smith, economist and author, The Theory of Moral Sentiment, believed that individuals, acting in their own self-interest, unintentionally benefit society as a whole. Greed, a more accurate dysphemism for self-interest, may motivate an owner to employ more people to make products, grow his operations and in the act of doing so, enables its employees to earn a living wage, be able to afford all that life has to offer, thereby adding wealth to companies benefitting from their purchases. These companies, in turn, will seek to expand labor opportunities as demand for their products increases. This domino effect is the unintended benevolent effect of the one individual acting in their own self interest.
Smith writes, “It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest. We address ourselves, not to their humanity but to their self-love, and never talk to them of our own necessities but of their advantages.”
Adam Smith tied self-interest to the market forces, and competition, which would drive down prices and curb exploitation.
AI’s Impact On Inequality
Joseph Stiglitz, Columbia University Professor and the Chief Economist of the Roosevelt Institute spoke at the recent International Association of Safety and Ethical AI Summit IASEAI in Paris, and he held that, “The pursuit of self-interest does not mean the well-being of society.” From his perspective, in the age of artificial intelligence, the interests of firms are not aligned with society.” In fact, the panacea that Smith described does not exist today. Among tech giants, self-interest is apparent in the time of accelerated artificial intelligence, but they are already operating in an economy where rising inequality has been intensifying for some time. As Stiglitz warns,
“There’s a lot of uncertainty about the speed of AI, but the risks are also increasing. Increasing inequality is one of the major problems our society faces today, which has contributed to societal polarization and dysfunction. I think that it is part of the root cause of what is likely to be the major global political dysfunction of the next four years of the current administration. However, the root cause is 45 years of growing inequality within the United States.”
He notes that the impacts are uncertain, but with certainty the net effect will be to reduce the demand for labor, increase inequality, increase unemployment, especially if the artificial intelligence materializes at full tilt.
The new tariffs only compound the problem. As per Chief U.S. Economist Ryan Sweet notes, “There’s nowhere to hide from tariffs.” They will disproportionately afflict the lower to middle income U.S. households, already dealing with inflationary burden.
What’s key is the asymmetric pace of change to the slower adaptation to technology. Stiglitz contends that if this pace of change is too rapid, “almost surely the adaptation of our society to this change will face problems reflected in unemployment and other societal dysfunctions.”
These dysfunctions have manifested themselves across the labor spectrum. The laws have historically lagged technology. The World Economic Forum pointed to only 22% of nations have policies to manage AI’s labor market disruptions, risking systemic instability. The OECD Employment Outlook (2024) projects that 56% of workers in OECD countries will require significant reskilling by 2030 due to AI automation, with low-skilled roles at highest risk of displacement. A recent report in Societies studied the effects of students who “over rely” on AI dialogue systems. One of the studies revealed, “regular utilization of dialogue systems is linked to a decline in abilities of cognitive abilities, a diminished capacity for information retention.”
An MIT 2022 study on AI and automation noted there will be “adverse effects on labor, including polarization of employment, stagnant wage growth for the middle and low-skill workers, growing inequality, and a lack of good jobs.” The paper acknowledges that while advanced economies will create good jobs, they are “skeptical” that ongoing automation will be able to ensure the “creation of enough good jobs.” In a recent Pew study among AI experts, 19% expect American will have more jobs in the next 20 years, while 39% expected fewer.
The existing harms that AI has enabled will also continue to worsen. Pew Research canvassed experts on their views of changes that could occur by 2035 in society’s use of digital systems and revealed 79% were concerned that technology will be primarily driven by profit motives, and “power incentives in politics.” They expressed the “fear as new threats to rights as privacy becomes harder, if not impossible, to maintain. They cite surveillance advances, sophisticated bots embedded in civic spaces, the spread of deepfakes and disinformation, advanced facial recognition systems, and widening social and digital divides as looming threats.” Experts also foresee a rise in crimes, online harassment and diminished human agency and new challenges in security. This will compound the rising unemployment, increased poverty and “diminished human dignity.”
And with these accelerated dysfunctions, Stiglitz asserts, “The effects on global inequality may be even worse … because AI, very quickly, is likely to have adverse effects on raw labor, unskilled labor, an asset that is relatively abundant in developing countries. So, while it may be able to optimize natural resource use,” he emphasizes this also reduces demand for both raw labor and commodities, devaluing the very assets that poorer nations rely on adding, “AI could dramatically widen inequality both within and between countries, with potentially severe economic and social consequences.” Stiglitz’s assertion means AI’s economic benefits may flow to wealthy nations, leaving developing countries further behind., and a destabilized global economic system.
AI And The Rise Of Monopoly Power
During his presentation, Stiglitz identified a critical uncertainty: “The second major unknown is how AI-driven market power will interact with inequality — we don’t yet have definitive answers, but I’ll offer some hypotheses.”
He questions whether we are witnessing a commodification, which assumes AI technology will be comparative to its rivals, largely compete on price, and be less subject to monopoly risks. Stiglitz reflects, “I am skeptical whether markets are rational or wise. Instead, current valuations may reflect hype rather than fundamental analysis. However, if one does trust the signals sent by the market, they are implicitly predicting a world where AI becomes a monopoly,” argues Stiglitz. He points to NVIDIA’s $2.7 trillion valuation, its dominance with 88% market share in the GPU market which suggests monopolistic forces are at play.
Instead, the existence of a trillion-dollar company, Stiglitz emphasizes, is a function of failure — not a success, “Our antitrust laws are failing, and as a result, our economy is no longer truly competitive. This has serious consequences — not just economically, but politically and, as he states, “When market power becomes highly concentrated in the hands of a few dominant firms, that economic influence translates into political power.” These corporations can then shape regulations, block antitrust enforcement, and prevent policies that would curb monopolistic behavior.
The Federal Reserve tracked the income and wealth distribution in the U.S. between 1990 and 2024. It revealed the highest earning Americans (1.0 % of population) have seen their share of total wealth grow from 16.6% (1990) to 25.7% (2024). Within this group, the top 0.1% share of wealth over the same 24-year period was more substantial from 8.6% to 13.8%.
According to Inequality.org, of the 0.1% wealthiest individuals, in the U.S. the top U.S. 5 billionaires own the biggest tech companies. 1) Elon Musk of Tesla/X and SpaceX with $428 billion 2) Jeff Bezos of Amazon with $235.2 billion 3) Larry Ellison of Oracle with $210.5 billion 4) Mark Zuckerberg of Meta with $204.4 billion 5) Larry Page of Google, with $157.6 billion. The billionaires’ club collectively grew their wealth by one trillion dollars over the last 9 months.
If Stiglitz’s assertion holds true, this tech billionaire’s club, could yield the same political power to weaken political oversight, which will lead to a self-reinforcing cycle of unchecked market power which will lead to unchecked political influence, and an entrenched monopoly.
Stiglitz again raises the question, “Whether there is monopoly power or not comes down to whether you can constrain market power and whether we are plateauing in AI (as stated earlier) or we are moving at an exponential rate? If AI moves at a much rapid pace, Stiglitz claims one company — or a small group — could gain decisive competitive advantage. They could use that advantage to dominate the market and effectively crush their competitors.
The grave outcome is the exacerbation of inequality as market power surges. This dynamic creates a major economic risk and with the pace at which AI progresses, the more likely that a winner-takes-all market emerges, as per Stiglitz. In this scenario, the winner can then exploit its power to control pricing, labor, innovation, thereby worsening inequality and reinforcing wealth concentration at the top.
Can economic self-interest in an accelerated development of Artificial intelligence ultimately yield societal benefits? From Stiglitz’ perspective, unless the interests of top AI firms align with society, this is unlikely. The ‘aligning’ mechanisms are regulatory. Today big tech is aggressively challenging long-standing legal protections in copyright law, in intellectual property, fair data practices, and data privacy protections. And while the current U.S. administration has effectively deregulated tech, mountains of criticisms persist in the name of public safety and safeguarding fundamental human rights.
If the optimal progress and benefit to society is to be realized, then self-interest needs to be checked with equalizing policy that limits the chances of any one entity being both the creator and arbiter.