AI is reshaping Wall Street! Goldman Sachs: Banks are facing automation adjustments like "human assembly lines"

According to John Waldron, President and Chief Operating Officer of Goldman Sachs Group, Inc. (GS.US), the bank is using artificial intelligence to expand its scale without the need to hire additional employees. "I often compare Goldman Sachs Group, Inc. to an artificial assembly line," Waldron said in an interview on May 12, throwing out this impressive metaphor. He then added, "Think about the development of the manufacturing industry, it has become more and more mechanized, more and more automated. The banking industry lags far behind in this respect." Behind these words reflects a silent revolution that a Wall Street giant with a history of over 150 years is undergoing. In this wave of technology driven by generative artificial intelligence, Goldman Sachs Group, Inc. is trying to turn "digital agents" into their "Siasun Robot & Automation", redefining the way the banking industry produces through AI. "Our human assembly line will become more digitalized, digital agents will become our Siasun Robot & Automation," Waldron said. "I am not sure how the overall number of employees will dynamically change, but I believe that the company will become more resilient and more scalable." Goldman Sachs Group, Inc. OneGS 3.0: A "AI-first" strategic transformation Waldron's remarks are not a passing whim. This popular candidate widely seen as succeeding David Solomon as CEO had previously introduced a strategic blueprint for Goldman Sachs Group, Inc. called "OneGS 3.0", aiming to systematically use artificial intelligence to improve efficiency and save costs. On April 29, 2026, Goldman Sachs Group, Inc. held its annual shareholder meeting in Salt Lake City, where all 13 directors, including Solomon and Waldron, were re-elected. The deployment of OneGS 3.0 was officially launched at the meeting, marking Waldron's AI strategy as a core agenda for the entire company. He joined the Goldman Sachs Group, Inc. board in 2025, becoming the second member to enter the board after Solomon, further consolidating his status as the successor. "OneGS 3.0" represents a multi-year transformation project, with its core goal being to embed AI as a foundational operational capability, rather than an isolated technical tool. The plan emphasizes process simplification, scalable productivity enhancements, and relies on high-quality data, shared platforms, and modern infrastructure to support the reliable operation of AI. In terms of specific areas of benefit, the strategy covers areas such as customer access, loan processes, regulatory reporting, and supplier management. These areas have long relied on a large amount of manpower for repetitive and rule-based operations, making them ideal directions for AI automation. Prior to this, Goldman Sachs Group, Inc. had introduced a generative AI assistant platform to about 46,000 employees across the company, significantly improving work efficiency in four key areas: quickly answering complex technical questions, summarizing the key points of dense documents, editing and improving written work, and brainstorming. A Goldman Sachs Group, Inc. partner revealed that she uses AI tools to handle up to 10 tasks a day, saving several hours of work time per week. Meanwhile, in early May 2026, Goldman Sachs Group, Inc. announced a $1.5 billion investment plan in partnership with Anthropic to accelerate the deployment of AI applications in its portfolio companies, deploying advanced models like Claude to improve efficiency and reduce costs. In measuring the effectiveness of AI deployment, Waldron revealed at the Goldman Sachs Group, Inc. RIA Professional Investors Forum that the bank evaluates success through three core dimensions: productivity gains, cost savings, and abandoned investments. The term "abandoned investments" refers to the efficiency gains brought about by AI that no longer require human and system input - a particularly intriguing metric, as it measures the hidden benefits of "things not done." Goldman Sachs Group, Inc. has set clear goals: to achieve a 15% efficiency improvement in the next three years, without massive layoffs. The bank also aims for a 60% efficiency ratio target, aiming to reduce costs to 60 cents for every dollar of revenue generated, which is a fairly ambitious level among large global investment banks. Data also shows that Goldman Sachs Group, Inc. employees save approximately 60 minutes a day on average using AI tools, which is considered to be the most specific AI productivity quantification metric on Wall Street to date. Layoffs and AI However, beneath the aura of efficiency improvement narratives, there is a looming cloud of layoffs on Wall Street. Since 2026, Goldman Sachs Group, Inc. has shifted towards a "rolling" layoff strategy, replacing the previous pattern of large-scale layoffs. The first round of layoffs began in April and will continue through the summer, affecting various business lines such as investment banking, asset management, and wealth management, but on a significantly smaller scale compared to previous years. Previously, Goldman Sachs Group, Inc. would typically cut up to 5% of its global workforce through annual strategic resource assessments (SRA), but in 2026, the bank opted for a more flexible, continuous small-scale adjustment path. However, Waldron provided an important clarification on this matter: many of the reported layoffs are not directly caused by generative AI, but rather a continuation of adjustments to the post-COVID-19 situation of "hoarding manpower" by companies. During the pandemic, the finance and tech industries experienced widespread recruitment expansions, and the current layoffs are more of a correction to this excessive expansion. True organizational change driven by generative AI, in Waldron's view, may not be more broadly visible until 2027 and 2028. "The cost savings brought about by generative AI may begin to have a more widespread impact on organizational structures in 2027 and 2028," he predicted in the interview. This assessment was also echoed in the reposts in Chinese media: "Many of the current layoffs are not directly driven by generative artificial intelligence, but are adjustments to the 'hoarding manpower' situation by companies post-pandemic." This timeline prediction resonates with current macroeconomic data. In the first quarter of 2026, layoffs in the U.S. tech industry exceeded 50,000, totaling over 80,000 for the year, but research from several institutions indicates that the impact of AI on the overall job market is still in its early stages. Morgan Stanley's research states, "The impact of AI on the labor market has been mild, with little evidence of widespread job losses." An analysis released by the Yale Budget Lab in April 2026 found that since the launch of ChatGPT, AI has not led to significant accelerated upheaval in the U.S. job market. However, warning signals are still strong. Goldman Sachs Group, Inc.'s research team estimated that AI had reduced monthly wage employment growth in the U.S. by about 16,000 jobs over the past year and pushed up the unemployment rate by 0.1 percentage points. The bank also issued warnings to its wealth management clients: tech industry employees replaced by AI may face significantly extended reemployment cycles and potentially lower salaries in new positions. "Diamond" or "Pyramid"? The unknown changes in organizational structure How will AI reshape the internal framework of institutions like Goldman Sachs Group, Inc.? Waldron's answer is: we do not know yet. "We don't know whether the future will be a 'diamond structure' or a 'pyramid structure'," Waldron said. This concise metaphor touches on a deep question: from which level will AI's penetration begin to change the organizational form of companies? Traditional companies have a pyramid structure - a vast base of junior employees supporting a narrowing middle to senior management levels. However, some industry insiders believe that AI technology may help streamline base-level employees, thus narrowing the base of the pyramid. If AI can take over the work of many junior analysts and back-office operations staff, companies may evolve into a "diamond" structure with a wide middle and narrow ends - where middle managers become the main body, while the scale of top strategic decision-makers and bottom-line operators is relatively limited. This evolution of organizational form will profoundly affect the career paths of future financial professionals. Goldman Sachs Group, Inc.'s approach provides an observation window: in 2025, while advancing AI automation, the bank promoted 638 bankers to Managing Director level, showing continued investment in high-end human capital. However, at the same time, the bank is also testing and preparing to introduce Cognition company's "AI engineer" Devin, planning to deploy hundreds to thousands of AI programmers in the future to handle tasks like legacy code maintenance. The productivity of a qualified developer using tools like Claude has increased several times. AI competition on Wall Street: Not just Goldman Sachs Group, Inc. Goldman Sachs Group, Inc.'s AI transformation is not an isolated case. The entire Wall Street is undergoing an "AI arms race." J.P. Morgan annually invests about $18 billion in technology, and CEO Jamie Dimon stated that $2 billion of AI investment has led to equivalent cost savings. The bank's CFO Marianne Lake indicated that AI has increased the bank's productivity impact from 3% to 6% to even higher levels, with some operational officers' productivity increasing by 40% to 50%. Bank of America Corp invests around $40 billion of its approximately $130 billion technology budget in AI and related new technologies, with its long-running virtual assistant Erica continuously driving front-end efficiency improvements. At the same time, the wave of layoffs by tech giants is also confirming the accelerated replacement effects of AI. In 2026, Meta announced a 10% layoff (about 8,000 people) and the closure of 6,000 vacant positions in recruitment; Microsoft Corporation launched its first voluntary retirement program in the company's 51-year history. According to Layoffs.fyi statistics, there have been over 81,000 tech professionals laid off since 2026, surpassing half of the total number for the entire year of 2025. The common feature of these layoffs is reallocating saved labor costs to AI infrastructure investments. However, even for Goldman Sachs Group, Inc., which is pushing AI the most, there are cautious voices within the organization. Goldman Sachs Group, Inc. partner Kerry Blum bluntly stated when discussing the limitations of AI tools: "The most significant limitation of AI tools may be the risk of overreliance. We must acknowledge that it is a tool, not a source of truth." She emphasized that in the banking business where clients are charged millions of dollars in fees, the ultimate responsibility for decisions still lies with people.