Economists Warn of AI’s Growing Impact on White-Collar Jobs

JPMorgan Chase and Goldman Sachs are leveraging AI to streamline operations, potentially leading to workforce reductions. Ford CEO Jim Farley has even suggested the technology could “replace literally half of all white-collar workers.” Salesforce’s Marc Benioff claims AI is already handling a significant portion – up to 50% – of the company’s workload. Walmart CEO Doug McMillon echoed this sentiment, stating AI “is going to change literally every job.”

The “it” driving these pronouncements across corporate America is, of course, artificial intelligence.

Less than three years into the generative AI boom, executives across major industries are openly discussing the potential impact on their workforces. The size and structure of these organizations are poised for dramatic changes, with many already underway.

What began with OpenAI’s ChatGPT and its consumer-facing chatbot capabilities has rapidly evolved into enterprise solutions. Companies are deploying customized AI agents to automate tasks in customer support, marketing, coding, content creation, and various other domains.

Goldman Sachs estimates suggest that 6% to 7% of U.S. workers could face job displacement due to AI adoption. The Stanford Digital Economy Lab, analyzing ADP employment data, has observed a 13% decline in entry-level hiring for “AI exposed jobs” since the proliferation of large language models. Their report identifies software development, customer service, and clerical roles as being particularly vulnerable in the current landscape.

“We are at the beginning of a multi-decade progress development that will have a major impact on the labor market,” said Gad Levanon, chief economist at the Burning Glass Institute, a research firm that focuses on changes in the economy and workforce.

Automation, in its various forms, is a recurring theme in economic history. From the printing press to ATMs and self-checkout systems, technological advancements have consistently reshaped the labor market. While some jobs become obsolete, new opportunities emerge, leading to economic adaptation and evolution.

A report from the World Economic Forum projects that AI, robotics, and automation could displace 92 million jobs by 2030. However, the report also anticipates the creation of 170 million new roles, particularly in areas such as AI development, research, safety, implementation, and robotics.

Erik Brynjolfsson, director of the Stanford research group, highlights the potential for new job creation and notes that physical roles such as health aides and construction workers are currently less susceptible to AI disruption.

“There’s going to be more turbulence in both directions in the coming months and years,” Brynjolfsson said in an interview. “We need to prepare our workforce.”

Despite the anxieties, macroeconomic data has yet to fully reflect a widespread impact.

While the U.S. government shutdown has temporarily halted the release of official labor statistics, alternative indicators, such as those from the Chicago Fed, paint a picture of moderate economic growth and a relatively stable labor market.

The Chicago Fed reported that the unemployment rate remained steady at 4.3% in September, with layoffs and separations holding at 2.1%.

A study published by the Budget Lab at Yale found no “discernible disruption” directly attributable to ChatGPT. Martha Gimbel, co-founder of the lab, describes the impact of AI as “minimal” and “incredibly concentrated,” acknowledging that the broader economic impact may unfold over time.

“The rest of the economy often moves more slowly than Silicon Valley,” she said.

Similarly, a survey by the New York Fed revealed that only 1% of services firms reported AI-related layoffs in the past six months. Data from the Society for Human Resource Management indicates that 6% of U.S. jobs have been automated by 50% or more, with that figure rising to 32% in computer and math-related occupations. These figures suggest a more nuanced reality than the often-hyped predictions of mass unemployment.

Corporate executives are increasingly vocal about their plans for AI implementation.

Amazon CEO Andy Jassy signaled in June that AI could lead to a reduction in the company’s corporate workforce over the next few years, encouraging employees to embrace AI tools to achieve “more with scrappier teams.” This shift towards more efficient and agile operational models is being examined closely by many companies.

The New York Times recently published an investigative report detailing Amazon’s efforts to automate its warehouse operations. The report, based on interviews and internal documents, suggests that automation could allow Amazon to avoid hiring over 160,000 people in the U.S. by 2027, resulting in significant cost savings.

In response to the Times’ report, an Amazon spokesperson stated the documents presented an “incomplete and misleading picture of our plans,” emphasizing that the materials reflected the perspective of one team and did not represent the company’s overall hiring strategy.

Palantir CEO Alex Karp has articulated a vision to increase revenue tenfold while simultaneously reducing headcount by approximately 12%. While he did not specify the timeframe for this ambitious goal, it underscores the company’s commitment to leveraging AI to enhance efficiency and productivity.

This message is gaining traction across the tech industry.

Salesforce CEO Benioff has disclosed that the company has reduced its customer support roles from 9,000 to 5,000, citing a reduced need for personnel due to automation. Swedish fintech firm Klarna reported a 40% workforce reduction as a result of AI adoption. Shopify CEO Tobi Lutke has challenged employees to demonstrate why AI cannot perform their tasks before requesting additional resources or headcount, emphasizing the importance of exploring AI-driven solutions across all functions.

AI-powered coding assistants have emerged as a key early application of generative AI, attracting significant user adoption and investment. Anysphere, the parent company of coding assistant Cursor, is reportedly in talks to raise funds at a valuation of $27 billion, positioning itself as a competitor to Microsoft’s GitHub and other players in the software development space. The increasing adoption of AI is not only lowering barriers to entry to software development, but also potentially creating code that has hidden risks, security concerns, or bias baked in.

However, software development is only the beginning.

In the banking sector, JPMorgan Chase has reportedly instructed its managers to limit hiring as it integrates AI across its various business units, according to CFO Jeremy Barnum. Goldman Sachs CEO David Solomon emphasized that AI integration will involve a comprehensive review of the bank’s organizational structure, decision-making processes, and overall productivity and efficiency.

The automotive industry is also deeply affected.

Ford CEO Farley’s remarks about AI’s potential to displace white-collar workers reflect a growing sentiment throughout the industry. A survey of U.S. car dealers conducted by Phyron suggests that half of the respondents expect AI to be capable of autonomously selling vehicles by 2027.

“That means AI creating the marketing assets, handling listings, answering buyer questions, negotiating deals, arranging finance, and completing the sale — all without human input,” Phyron explained in its survey report.

The impact of AI will likely be a prominent topic during the upcoming earnings season, as major tech companies provide updates on their AI initiatives. Tesla will kick off tech earnings on Wednesday, followed by Alphabet, Meta, Microsoft, Apple, and Amazon the following week. Investors will be scrutinizing these reports for insights into the strategic direction of these firms, their AI investment plans, and, crucially, their assessment of AI’s potential impact on the workforce.

The long-term economic effects hinge on the balance between job displacement and new job creation. While certain sectors may experience workforce reductions due to automation, entirely new industries and roles are likely to emerge in fields related to AI development, implementation, and maintenance. The key question for policymakers and educators is how to best prepare the workforce for this rapidly evolving landscape, ensuring that individuals possess the skills and knowledge necessary to thrive in an AI-driven economy.