Goldman Sachs Group Inc (NYSE:GS, XETRA:GOS) CEO David Solomon said fears of a broad “job apocalypse” driven by artificial intelligence are overblown, arguing instead that AI will reshape work rather than eliminate it.
In a New York Times opinion piece published over the weekend, Solomon argued that generative AI should be understood as part of a long historical pattern of technological change that ultimately expands economic opportunity even as it disrupts existing roles.
Solomon compared the current wave of AI development to earlier innovations such as electricity, the digital revolution, and spreadsheets, which he said displaced certain tasks but also enabled new categories of higher-value work.
He rejected what he described as “doomsayer” narratives about mass unemployment, framing the shift as a continuation of “creative destruction” rather than a structural collapse of employment.
Citing internal Goldman Sachs analysis, Solomon said AI could automate roughly 25% of current work hours over the next decade. However, he argued that this would not translate into equivalent job losses, but rather a reallocation of time toward more complex, client-facing, and strategic responsibilities.
The firm’s analysts also highlighted areas where AI is expected to reduce demand for certain types of labor, including data-heavy functions such as regulatory reporting and client onboarding. At the same time, they noted that AI has driven increased demand in other parts of the economy, including data center construction and related infrastructure, which the bank estimates has supported hundreds of thousands of US jobs since 2022.
That shift, Solomon said, is prompting firms to concentrate hiring in relationship-driven roles such as investment banking, trading, and asset management.
Under Solomon’s framework, the broader economic impact of AI is expected to be shaped by three main dynamics: the optimization of human work as routine tasks are automated, higher performance standards within existing roles, and the emergence of new positions focused on managing, auditing, and governing AI systems.
Despite this long-term optimism, Solomon’s article acknowledged that AI is already changing workforce composition within financial services. Entry-level and junior roles, in particular, are facing pressure as AI tools increasingly handle tasks such as financial modeling, note-taking, and spreadsheet analysis that traditionally formed the foundation of analyst work.