AI Unemployment Panic? Hard-Hitting Wall Street Research Report Strikes Back: Programmer Jobs Surge by 11%

Early in 2026, an AI doomsday report emerged out of nowhere, directly triggering a panic-driven crash in the U.S. software sector. Two hundred billion dollars in market value evaporated overnight, leaving countless white-collar workers anxious about being "replaced by AI tomorrow."

However, just as public panic intensified, a重磅 research report from Citadel Securities, the world's largest market maker, swiftly arrived. Armed with hard data, it shattered these "rumors":

AI is not a job-destroying monster but a core force countering headwinds of the times. Even software engineering roles, considered most vulnerable to replacement, have surged by 11% year-over-year.

This AI-era employment anxiety is merely an amplified illusion.

A Report Triggering a Trillion-Dollar Market Earthquake: Why Did the AI Doomsday Narrative Resonate?

On February 22 (U.S. local time), Citrini Research released its report titled "2028 Global Intelligence Crisis."

From a futuristic perspective, it painted a dystopian picture driven by AI: mass white-collar unemployment, consumption collapse, financial system failures, and iterative AI advancements triggering cascading unemployment spirals.

This report acted like a bomb dropped into the market, causing a collective stampede in the software sector the very next day.

The iShares Software ETF (IGV) plummeted by as much as 4.8%. Since early 2026, stock prices of software giants including Salesforce, Workday, and Snowflake have collectively crashed between 20% to 50%. Market fear regarding AI replacing jobs reached its peak.

After all, in public perception, AI's rapid evolution inevitably means the demise of human jobs, especially for software engineers who write code, seemingly destined to be the first group replaced by AI.

But is this really the case?

Frank Flight, a macro strategist at Citadel Securities, provided the most direct answer with a rebuttal report titled "2026 Global Intelligence Crisis": NO!

Source URL: https://www.citadelsecurities.com/news-and-insights/2026-global-intelligence-crisis/

Hard Data Shatters Unemployment Myths: Programmer Jobs See a V-Shaped Reversal

Every dataset in Citadel's report颠覆 s public perception, causing the AI unemployment theory to collapse under its own weight.

Software engineers, seemingly cornered by AI, are experiencing a beautiful V-shaped recovery in the recruitment market:

Since bottoming out in May 2025, the number of software engineer positions has climbed steadily, surging 11% year-over-year. Recruitment data from the Indeed platform clearly shows that demand for software engineering roles is far outpacing the overall hiring market, with热度 remaining exceptionally high.

Amidst significant AI development, the overall U.S. unemployment rate has not skyrocketed; instead, it has remained steadily within a reasonable range at 4.28%.

Furthermore, capital expenditure on AI has reached $650 billion, accounting for 2% of the nation's GDP. In 2026, the U.S. plans to construct 2,800 data centers. Commodities related to the AI supply chain have surged by 65% since January 2023. This vigorous AI infrastructure boom has not created unemployment; on the contrary, it is generating more job opportunities.

On one side lies the panic of hundreds of billions in evaporated market value; on the other, the reality of surging jobs and stable unemployment. Behind this stark contrast hides a core cognitive misconception about AI: equating the recursive potential of AI technology with the recursive deployment of the economy.

Simply put, while AI can indeed self-evolve and become smarter, enterprises simply cannot deploy AI into every aspect of daily work at an exponential speed.

Real-time population survey data from the Federal Reserve Bank of St. Louis further poured cold water on the AI replacement theory. This survey does not ask whether AI is used, but focuses on the frequency of AI usage in the workplace. The results show that the proportion of AI usage in the workplace has remained stable, without any steep upward inflection point. This indicates that the popularization of AI in daily work has not yet reached the so-called "acceleration phase."

Let us now examine recruitment demands within China.

Since March, technical roles have accounted for 85% of Ant Group's recruitment drives. These specifically include AI research, AI applications, AI infrastructure, and other areas, with a strong focus on core fields such as large model algorithms, multimodal generation, data intelligence and basic platform R&D, embodied AI, and AI security.

Other leading domestic tech giants have also launched a fierce "talent war" for AI. Both Tencent and ByteDance announced the launch of their 2026 internship recruitment programs, with a combined total of over 17,000 positions available. According to Tencent, this recruitment drive expands technical roles by 36% and product roles by 39%, with a significant increase in AI-related positions. In ByteDance's current recruitment cycle, over 4,800 offers for R&D roles are pending issuance, marking the highest number in the company's history.

AI Adoption Is Never Overnight: Physical Boundaries and the S-Curve Ensure It Remains a Human Assistant

The root cause of excessive public anxiety regarding AI replacement lies in ignoring the objective laws of technology adoption and underestimating the realistic constraints of AI implementation.

Historically, the adoption path of generative AI mirrors that of personal computers and the internet in years past, all following a long and tortuous S-curve: slow and costly early adoption, gradual acceleration only after infrastructure matures, and finally entering a plateau phase once the market saturates.

Today's AI is still in the early stages of the S-curve, far from comprehensively replacing humans. More importantly, for AI to fully replace humans, it faces an insurmountable physical boundary.

Training and running AI requires massive semiconductor capacity, enormous data centers, and staggering amounts of electrical energy. If AI were to take over all white-collar jobs, the required computing power would be several orders of magnitude higher than current levels. Once the demand for automation surges, the marginal cost of computing power will climb sharply. When the cost of machine computing exceeds human wages, the so-called "comprehensive replacement" becomes an uneconomical proposition.

Beyond this, AI implementation is constrained by chip supply, grid capacity, regulatory approvals, and cumbersome organizational changes within enterprises. It can never be replicated and promoted "frictionlessly." No matter how advanced the technology, it must ultimately take root in the soil of the real world.

Looking Back at History: Technological Progress Has Never Eliminated Jobs; Instead, It Catalyzes New Opportunities

When discussing technology replacing labor, Keynes' prediction of a "15-hour workweek" is an unavoidable classic.

In 1930, this economist predicted that with a substantial rise in productivity, humans would only need to work 15 hours a week by the early 21st century. He was right about the first half regarding productivity gains but misjudged the labor market outcome. Human desire possesses infinite elasticity. As productivity rises and lowers commodity costs, people did not choose to lie flat. Instead, they used their increased wealth to pursue higher-quality goods and entirely new services, leading to the continuous birth of new job opportunities.

Citadel's report explicitly points out that AI-driven automation is essentially a positive supply shock. Its mechanism is identical to that of the steam engine, electrification, and computers: it does not destroy aggregate demand; instead, it lowers marginal costs and enhances society's actual purchasing power. As enterprises produce more products at lower costs, commodity prices fall, residents' real incomes rise, and a plethora of new demands and industries emerge accordingly.

Data from the U.S. Census Bureau corroborates this point: the number of new business formations in the U.S. is currently growing rapidly. Behind this entrepreneurial boom lie new tracks and commercial opportunities catalyzed by AI. Meanwhile, various leading indicators in the labor market also show a trend of comprehensive improvement.

To build the planned 2,800 data centers, recruitment demand in the U.S. construction sector has rebounded strongly. Gaps for electricians, construction workers, network technicians, and other roles continue to widen. As NVIDIA CEO Jensen Huang stated, AI is the largest infrastructure project in human history. While creating economic returns in energy and semiconductors, it is also generating a vast number of new jobs in infrastructure, operations, and maintenance.

Just as people once worried that office clerks would be completely unemployed when Microsoft Office was born, it ultimately just became an efficient auxiliary tool for humans.

The same applies to AI today. In fields requiring high levels of physical coordination, complex interpersonal relationship handling, strict regulatory review, and deep trust barriers, the role of humans remains irreplaceable. What AI can do is liberate human hands, allowing us to step away from repetitive tasks to engage in more creative and meaningful work.

AI Is Not the End of Days, But a Glimmer of Hope Hedging Against Headwinds of the Era

For AI to trigger a severe economic depression, a series of unrealistic conditions would need to be met: instantaneous peak technology adoption, 100% replacement of human labor, zero fiscal hedging by governments, cessation of corporate profit reinvestment, and infinite expansion of computing power disregarding physical laws. In reality, once faced with extreme replacement risks, social institutions will inevitably hedge through regulatory and fiscal policies. This is the inevitable logic of human societal development.

Looking back at technological transformations over the past century: automobiles replaced horse-drawn carriages but gave rise to drivers, auto repairmen, and traffic management roles; telephones replaced messengers but spawned careers in telecommunications operations, customer service, and communication technology; the internet disrupted traditional industries but created entirely new tracks like e-commerce, live streaming, and big data.

No technology has ever truly rendered human labor redundant. On the contrary, every technological revolution has pushed human society toward higher quality development.

The current world faces three major headwinds of the era: an aging population, climate change, and de-globalization. Layered together, they act as "debuffs" on global economic growth. The productivity boost brought by AI is precisely the key force to counter these downward pressures.

AI is not an "employment doomsday" that smashes rice bowls; it is an effective strategy to help us 对抗 era challenges and open up entirely new development spaces.

Those panics about AI unemployment will eventually dissipate as technology lands and cognition improves.

Rather than immersing ourselves in anxiety about being replaced, it is better to actively embrace AI and learn to coexist with this new tool. After all, in the long river of technological development, the true crisis has never been the technology itself, but the mindset of refusing to change and remaining stagnant. In the AI era, the true "iron rice bowl" is never a specific job title, but the ability to continuously learn and evolve.

Author: Kaigong de Daxiong

Selected Articles: