Goldman Sachs is pushing deeper into synthetic intelligence — and what it may imply for the agency’s future.
In its 2025 shareholder letter, launched Friday, Goldman’s leaders doubled down on their priorities tied to the recent refresh of their “One Goldman Sachs” initiative to streamline the financial institution’s providers and enhance income.
The agency described One GS as a “new working mannequin propelled by AI,” and highlighted six areas “ripe for disruption”: consumer onboarding and KYC, vendor administration, regulatory reporting, lending, enterprise danger administration, and gross sales enablement.
“This does not simply imply retooling our platforms,” the financial institution mentioned. “It means taking a front-to-back view of how we set up our individuals, make selections, and take into consideration productiveness, effectivity, and resilience.”
Goldman signaled that executing on its AI ambitions will rely closely on expertise, the place a battle continues to unfold to nab top performers throughout Wall Road.
“Competitors from inside the monetary providers business and from companies exterior the monetary providers business, together with the know-how business, for certified workers has typically been intense,” the agency mentioned. It added that it has “skilled elevated competitors in hiring and retaining workers” tied to its know-how initiatives and newer enterprise strains.
That stress is especially acute in newer hubs central to Goldman’s working mannequin. The financial institution mentioned 45% of its workforce is now primarily based in strategic areas like Warsaw, Bengaluru, Hyderabad, and Salt Lake Metropolis. In these markets, Goldman typically competes with corporations which have a deeper native presence.
On the similar time, the agency continues to attract important curiosity from candidates. Goldman mentioned it acquired a couple of million skilled rent functions in 2025, up 33% from the prior yr.
The most important dangers
Along with shedding some mild on its future AI bets, Goldman additionally defined what it sees as the largest dangers tied to its use of the know-how that it believes shareholders must know.
The agency mentioned “the authorized and regulatory atmosphere referring to AI is unsure and quickly evolving,” and warned that AI fashions might generate incorrect outputs.
Generative AI models, the agency flagged, are susceptible to creating errors, which may, in a worst-case situation, “end result within the launch of personal, confidential or proprietary data, that mirror biases included within the knowledge” they’re educated on. And it pointed to its use of “AI fashions developed by third events” which the agency mentioned makes it “dependent” on how these suppliers construct their fashions. What’s extra, unhealthy actors may harness AI’s capabilities “to commit fraud and misappropriate funds and to facilitate cyber assaults,” it mentioned.
Nonetheless, CEO David Solomon struck a bullish tone within the letter about his prognosis for AI. “We consider this know-how goes to reshape the best way we dwell and work,” he wrote, including that “on the similar time, there are important questions” concerning the mind-bending velocity of its adoption.
“With any new know-how, there shall be winners and losers,” the CEO added. “Whereas there are prone to be intervals of recalibration, in the long term I consider the online advantages from AI will accrue to many establishments as AI funding continues to construct.”
AI has been a central focus for Solomon. The financial institution has partnered with builders like Cognition Labs to create distinctive merchandise and rolled out its GS AI chatbot to its greater than 47,000 workers.
Talking at a convention in Europe final yr, he mentioned he wished Goldman may spend greater than its roughly $6 billion technology budget, however famous on the time that funding ranges have been considerably constrained by the necessity to ship shareholder returns.
