Goldman Sachs CEO says AI will help companies hold staffing flat

David Solomon is putting AI in the language of staffing control, not just productivity. In a podcast interview recorded in December and published April 15, the Goldman Sachs chief executive said large companies are trying to keep head count flat while using artificial intelligence to improve operating efficiency, a sharper message than the usual talk about faster workflows.

Solomon framed the macro backdrop as supportive, citing AI investment, easier regulation and persistent fiscal stimulus. But his clearest point was about management discipline. Companies, he said, are increasingly asking whether they can “keep head count from drifting down by 2%, 3%, or 4%” and use AI capability to preserve staffing levels while pulling more output from the same organization. For Goldman employees, that is not a vague innovation slogan. It is a sign that the bank’s leadership sees AI as a tool for tightening hiring, backfill and performance expectations.

That matters inside a firm that still depends on human judgment, apprenticeship and long hours to produce revenue. Goldman reported 47,400 employees as of Dec. 31, 2025, up from 46,500 in 2024, even as the firm kept pressing for better efficiency and higher returns. In 2025, Goldman said net revenues rose 9% to $58.3 billion, earnings per share reached $51.32 and return on equity hit 15.0%. The bank also said it had reduced historical principal investments by more than 90%, from roughly $64 billion to $6 billion, underscoring how far it has moved toward a more capital-light model.

The staffing message also fits a longer pattern. Goldman’s 2024 annual report said its efficiency ratio improved to 63.1% and total shareholder return was 52%, evidence that management has been rewarding leaner execution. In October 2025, Reuters reported that Goldman had told employees about potential job cuts and a hiring slowdown tied to an AI initiative called OneGS 3.0. Reuters later reported in March 2026 that the firm planned to cut a small number of underperforming staff in April as part of its strategic resource assessment, which traditionally affects 1% to 3% of staff.

Solomon’s remarks suggest AI is now part of Goldman’s workforce management strategy, not just a client pitch. For analysts and associates, that likely means fewer automatic backfills, more pressure to show measurable output, and a higher premium on people who can use AI to increase throughput without adding layers. For a bank that still prizes prestige and promotion velocity, the risk is clear: if technology helps the firm hold staffing flat, the same work can be spread across fewer incremental roles, raising the bar for advancement even when the firm is growing.