In this discussion, we delve into how technological advancements, particularly in AI, serve as tools for extracting value from the workforce. The shift towards shareholder capitalism in the Anglosphere emphasizes prioritizing stockholders, even though equity should legally be the last claim on profits after other obligations are met. This concept of shareholder primacy, touted by economists, lacks contemporary legal backing, especially when examining guidance from major law firms for corporate board members. It’s less about maximizing shareholder wealth and more about ensuring corporate survival.
By Lynn Parramore, Senior Research Analyst at the Institute for New Economic Thinking. Originally published at the Institute for New Economic Thinking website
Have you noticed how predictable discussions about AI have become?
Workers are urged to quickly adapt and become “AI-ready,” while influential business figures and entrepreneurs preach that embracing AI is not just beneficial, but a moral obligation. The promise is appealing: engage with AI, and you may find less mundane work, enhanced creativity, and perhaps reduced hours. However, if you hesitate, the forecast isn’t bright.
Yet, when analyzed through the lens of shareholder value, a different narrative emerges. Over the past four decades, increasing productivity has meant significant rewards for wealthy executives and major shareholders, while workers are left merely thankful for retaining their positions.
AI is being introduced into a framework designed to funnel wealth upward. Why would that change now?
Let’s take a look at the unfolding AI landscape.
For instance, consider Microsoft. The allocation of its profits can shift annually, but it’s a safe bet that a substantial portion goes to shareholders and top executives whenever the company sees financial gains.
Between 2006 and 2025, Microsoft allocated $283 billion to stock buybacks and $223 billion to dividends — together making up 71% of its profits. Even as Microsoft invests heavily in AI, it also sends large amounts to shareholders while laying off thousands of employees, leaving those who drive the company’s success uncertain about their incomes.
CEO Satya Nadella doesn’t share that concern.
In 2025, he received around $123 million, more than double his previous year’s compensation. This astonishing sum, as outlined by William Lazonick and Matt Hopkins, reflects actual money executives take home, contrasting with the often-reported estimates of stock awards. Over his 13 years as CEO, Nadella has pocketed nearly $1.2 billion — averaging $91 million annually. The bulk, approximately 90%, arises from stock-based compensation. In 2025, Microsoft’s board granted him over $84 million in stock — the highest of his tenure. While these shares might not provide immediate cash, if Nadella can sustain Microsoft’s stock price with hefty buybacks, his financial gains could expand significantly.
Nadella’s enormous compensation structure incentivizes other executives to keep Wall Street satisfied, often at the cost of employee welfare.
Now, consider Nvidia, another major player reaping the benefits of the AI surge. The spike in demand for its chips has led to an influx of cash that is invested in research, cutting-edge AI systems, and attracting top talent with hefty salaries. It sounds promising! However, shareholders expect immediate returns.
This results in Nvidia engaging in tens of billions of dollars in stock buybacks — with even more anticipated — despite competing in an undoubtedly resource-intensive sector. Lazonick notes that Nvidia’s recent buyback strategy was prompted by soaring profits, as founder and CEO Jensen Huang holds merely about 3% of the company’s stock. “He must appease hedge fund activists or risk losing his position,” Lazonick highlights.
This pressure raises significant concerns. Nvidia is attempting to innovate while simultaneously appeasing shareholders, who consistently demand more.
Who are these hedge fund activists influencing Huang? They are shareholders who purchase stakes in companies, then urge leadership to implement changes aimed at increasing stock prices. Their suggestions may involve replacing executives, trimming costs, or increasing shareholder payouts. These activists don’t always directly manage the company or necessarily understand how to improve it, but their stock holdings give them a powerful voice in corporate decisions. Lazonick elaborates that these individuals are not “investors” but rather “predatory value extractors” who benefit from the contributions of the true value creators: the workers.
At this moment, Nvidia executives believe they can finance the AI race while also buying back shares. However, looking at Intel provides a cautionary lesson: the company focused heavily on shareholder rewards while losing its technological edge. This framework is far from obsolete.
Even Nvidia’s well-compensated engineers are not immune to the logic of shareholder value; they simply occupy a higher tier in the expense structure. If financial pressures arise, they may find themselves at risk of hiring freezes, project eliminations, or the dreaded “strategic restructuring.” Yet, somehow, an avenue for another buyback always seems to be present.
Executive compensation operates on a separate trajectory. According to Lazonick, stock-based pay has a “ratchet effect,” setting new standards for future earnings packages. What once appeared excessive quickly becomes commonplace, leading to billionaire paychecks merely being the norm.
Moreover, startups that have yet to go public are not exempt from this dynamic. Companies like OpenAI and Anthropic may not engage in buybacks now, but they already confront the demands of the shareholder value machine: securing funding, pursuing astronomical valuations, and convincing shareholders that high expenditures are justified. The funds flowing into AI are often branded as “investments,” but Lazonick points out that this spending is primarily a gamble on who will ultimately profit and who will be left stranded when stock prices crash. Do these speculators genuinely care about ensuring companies achieve productivity and profitability? The answer appears to be no.
One might wager that if the current plight persists, AI will not yield benefits for most American workers.
Elon Musk’s ascent to trillionaire status vividly illustrates the workings of the shareholder value ideology. When Tesla and SpaceX gained significant value, who reaped the rewards?
Musk’s vast equity stake provides the answer. He acquired a large number of Tesla shares through compensation packages, which he argued were essential to safeguard the company from potential takeovers. As Tesla successfully ramped up electric vehicle production, coinciding with consumer demand, the rise in market valuation translated Musk’s shares into incredible personal wealth.
“Tesla’s significant shareholders increased their wealth with rising share prices,” Lazonick explains, “despite the fact that the company’s transformation was accomplished by engineers, factory workers, suppliers, and the individuals who designed, built, and scaled Tesla’s vehicles.”
SpaceX reflects similar trends. Musk maintained a significant equity position while establishing a high private-market valuation, thus accumulating substantial paper wealth as the valuation continued to soar. Once SpaceX becomes public, shareholders are likely to prioritize stock appreciation over enhancing the company’s productive capabilities.
Why innovate for the future when one can maximize profits from previous successes?
As Lazonick aptly put it, “Musk has mastered the Wall Street playbook — he embodies a founder with a massive equity stake, genuine innovation, and traders pursuing the stock, transforming technological progress into another machine for wealth concentration.”
The stakes are rising. Shareholders might choose to stay or leave, but this uncertainty underpins the entire Silicon Valley ecosystem. Companies promote exaggerated expectations about AI’s potential, and this hype can translate into billions before any actual profits materialize.
Economist Servaas Storm argues that CEOs touting AI are making “unhinged” and unverifiable claims about a future devoid of work or overflowing with abundance and productivity. In reality, many prominent AI players are merely redistributing funds, creating a precarious structure based on the assumption that technology will eventually fulfill promises. For instance, Microsoft supplies the cloud infrastructure for OpenAI while also purchasing OpenAI’s offerings. This interdependent cycle — money flowing from Microsoft to OpenAI and back again—allows both companies to showcase growth, albeit driven by intertwined transactions rather than an independent market.
If genuine revenue fails to materialize, it likely won’t be Sam Altman or Satya Nadella facing repercussions.
Conversely, workers are directly experiencing the ramifications of AI on their jobs: cuts in headcount, vanishing entry-level positions, increased surveillance, algorithm-determined schedules, and what Cory Doctorow has aptly termed a pervasive “enshittification” of employment conditions.
How dire is the situation? Let’s enumerate the effects.
Amazon has intensified warehouse and customer-service automation alongside stringent performance monitoring. IBM has publicly linked AI integration with reductions in back-office jobs. Microsoft and Google are integrating AI into all their products, claiming a revolution in productivity while silently assessing how many employees they can terminate. Furthermore, workers at Meta have accused the company of targeting those on maternity or disability leave for layoffs using AI. In response to workforce reductions, Google employees are demanding protections against layoffs.
Despite AI being capable of generating reports and codes in seconds, the projected gains in productivity rarely translate to increased leisure for workers — often yielding the opposite. Employers tend to announce, “Here are new tools — now work faster and harder. And don’t expect a raise.”
AI might be viewed as the ultimate technology for shareholder benefit — a tool to enhance productivity while further diminishing labor power, irrespective of stock market conditions. As Lazonick succinctly stated, “When stock prices drop, layoffs follow. Workers face pay cuts while executives receive significant stock bonuses. Then, when stock prices rise again, executives reap substantial rewards, disproportionately capitalizing on the upside.”
He warns that integrating AI into this flawed system could widen inequality to an unprecedented degree.
“It enriches a select few, and an even smaller elite become exceptionally wealthy, all revolving around stock values,” he explains. “Those accumulating extraordinary wealth feel entitled due to the deep-rooted ideology of shareholder value that persists unchallenged.”
Until the doctrine of shareholder value is openly contested, there’s minimal reason to assume that AI, despite its potential benefits, will meaningfully benefit the workers powering it. Profit will remain privatized while disruptions are socialized, echoing trends from the past decades.
The Shareholder Era Was Constructed, and It Can Be Dismantled.
How can we rewrite this narrative?
The initial move is recognizing that the shareholder value framework is a relatively recent construct.
Throughout much of the postwar years, Americans believed corporations bore responsibilities to society for the privileges granted to them. They were expected to produce valuable goods, ensure stable jobs, invest in the future, and share success with workers, communities, and taxpayers who contributed to their achievements. As Lazonick often highlights, the post-WWII economic boom in the U.S. resulted in upward mobility for most Americans, which contradicts any notion of shareholder primacy. Such norms shifted in the 1980s, but it’s possible to reclaim and reshape them.
The contrast in the earlier mindset became evident in places like Detroit’s auto plants and Midwestern steel mills. Increased productivity led to higher wages, pensions, and a genuine belief that each generation could enhance its living standards. Unions played a pivotal role in this, as did corporate practices that prioritized investment, workforce development, and shared benefits. The public did not anticipate that economic growth would concentrate wealth at the top. Why should we accept that now?
Lazonick illustrates, “Shareholder value faced limitations due to the influence of unions and managers who acknowledged the need for collaboration with unionized workers to create value. Despite having boards elected by shareholders, norms rendered the current exploitation of workers unacceptable.”
Fortunately, signs indicate shifting norms may be on the horizon. Unions are more popular than they have been since the 1950s, and an increasing number of Americans are expressing dissatisfaction with the disproportionate power of corporations.
The next step is debunking myths that have been accepted so broadly they seem factual.
Myth #1: Corporate law mandates that executives must maximize shareholder value. This is not true. As legal expert Lynn Stout and others have verified, corporate law allows for managerial discretion beyond merely focusing on stock prices. “If maximizing shareholder value was necessary, then managers must have been breaking the law for much of the 20th century,” Lazonick quips. “The post-WWII surge in the U.S. economy and societal advancement for the majority must be evidence of their wrongdoing.”
Myth #2: Shareholder value equates to sound business practice. Contrary to decades of evidence, buybacks, layoffs, and financial engineering may yield short-term benefits for equity holders but can weaken companies in the long run. A landscape fixated on shareholder gains leaves workers vulnerable and the wider economy less capable of translating productivity increases into communal prosperity. During the shareholder value era, U.S. corporations have increasingly trailed international competitors across various critical technologies.
Myth #3: This is simply how capitalism functions. This is incorrect. The push for maximizing shareholder value stemmed from specific alterations in corporate governance, financial markets, and public policies. Americans didn’t adopt this idea due to conscious debates; rather, it became established through laws, economic incentives, executive compensation structures, market dynamics, and academic institutions. These can all be challenged and reformed.
Even many proponents recognize that the system is not immutable. In 2019, the Business Roundtable — essentially the voice of the nation’s top executives — collectively stated that corporations should cater to workers, customers, suppliers, and communities, not solely to shareholders. This declaration wasn’t coincidental as Elizabeth Warren’s proposed Accountable Capitalism Act aimed to reform the shareholder value system by requiring large corporations to consider broader stakeholder perspectives and grant workers representation on corporate boards. Although the Roundtable’s statement may have been convenient public relations, it indicated an acknowledgment of shifting political climates.
Lastly, it’s essential to recognize that individual workers alone cannot tackle the challenges posed by AI.
Workers continually hear calls to reskill, adopt AI, and boost productivity. Yet, as Thomas Ferguson states, “We face a distribution problem, and no amount of reskilling can remedy a system where productivity gains are preordained for shareholders.”
This view reframes our understanding of technology. If AI amplifies productivity, we must question whether that translates into improved wages, shorter workweeks, enhanced job security, and increased investments in workers. Or, will it merely lead to another round of stock buybacks, extravagant executive paychecks, and more megayachts?
The outcome hinges on corporate governance, underscoring the need for reforming the rules and incentives that have facilitated the upward redistribution of wealth for the past four decades.
Here are several strategies to consider:
- End stock buybacks. Lazonick contends that stock buybacks should once again be classified as market manipulation, much as they were prior to SEC Rule 10b-18 in 1982, which relaxed restrictions. They should be prohibited entirely, not merely taxed.
- Rethink executive compensation. It’s crucial to incentivize executives for nurturing robust companies and extending productive capabilities rather than simply inflating stock prices. Compensation should reflect investments in workers, innovation, productive capacities, and the generation of sustainable value instead of payouts to shareholders and executives. High rewards ought to be reserved for consistent, long-term success rather than occasional spikes.
- Change the corporate governance landscape. Workers and taxpayers take on significant risks in corporate success or failure. They deserve notable representation on boards alongside shareholders who prioritize building productive firms.
- Empower workers. AI should not serve as a means of surveillance, deskilling, or avoiding union relationships. Workers require stronger negotiating rights and genuine participation in decisions regarding new technologies.
- Don’t confuse an AI dividend with economic reform. Plans like Bernie Sanders’ proposal for an AI fund acknowledge the public’s right to share in the wealth generated by AI. However, relying solely on a dividend risks addressing the symptoms without tackling the root issues. This allows corporations and political entities capturing the majority of these gains to continue business as usual. While a $1,000 check might provide temporary relief, it pales in comparison to the vast wealth already amassed at the top. Our priority must be ensuring that those who create wealth can share in it from the outset.
- Invest in people’s ability to harness AI. Although retraining is crucial, it’s even more vital to ensure individuals have access to education, skills, healthcare, security, and opportunities necessary for leveraging new technologies to improve their lives. Investments in workers, coupled with enhancements to educational, healthcare, and public institutions, are essential for all to partake in the benefits provided by AI.
- Limit corporate political spending and lobbying. Reducing large monetary influence in politics is paramount. AI companies shouldn’t wield their wealth to shape governing rules to their benefit. This may ultimately require a constitutional amendment to reverse legal rulings granting corporations extensive political spending rights. Corporations are legal entities, not individuals; they exist through societal privileges and should therefore prioritize the public good, rather than utilizing wealth to silence the voices of those affected by their actions.
Importantly, these proposals do not suggest placing profits off-limits or rendering government the true owner. They merely advocate for a reevaluation that once again balances the interests of those who create value against a singular focus on shareholders.
Addressing these issues politically is challenging; however, past victories like the eight-hour workday, legalizing collective bargaining, and establishing social protections were similarly arduous.
Norms shift when people realize that the narrative they’ve been sold does not align with reality. For decades, Americans were led to believe that prioritizing shareholders would enhance everyone’s well-being. Yet countless individuals have witnessed the outcomes of that doctrine — greater wealth concentration at the top while their economic circumstances stagnated or deteriorated. Over time, ideas initially viewed as radical may emerge as the necessary corrections.
If companies pledge to treat workers fairly, it signals an inclination toward corporate social responsibility. However, what is truly desired is for these corporations to advocate for regulations that uplift everyone: stronger collective bargaining rights, greater worker representation on corporate boards, limitations on workplace surveillance technologies, modified tax and corporate regulations governing buybacks and executive compensation, more stringent antitrust enforcement, and enhanced public investment in worker training and education. Additionally, we must quell corporations’ attempts to use political contributions to counterbalance reforms meant to enforce accountability. This illustrates what genuine transformation entails.
We need not allow the AI era to become yet another mechanism of wealth transfer. With proper governance, corporations can resume their role of serving those who contribute to their success, fostering prosperity for all Americans. Other successful economies manage this, and so did the United States in the not-too-distant past.
Corporations were established to benefit the public. Somewhere along the line, we lost sight of whom they are meant to serve: us.
