Goldman Sachs Partners With Anthropic To Deploy AI Agents

Goldman Sachs has partnered with AI startup Anthropic to deploy autonomous AI agents, described as digital co-workers, across its investment bank. The firm spent six months co-developing these tools with embedded Anthropic engineers using the Claude model to automate process-intensive back-office functions. Targeted operations include trade and transaction accounting, regulatory compliance checks, financial statement reconciliation, audit preparation, internal reporting, and client onboarding and due diligence.

Chief Information Officer Marco Argenti stated the technology is designed to reason through complex problems and significantly reduce the time required for essential functions. The initiative is part of a broader multiyear reorganization led by CEO David Solomon to center the bank's operations around generative AI while constraining headcount growth. Argenti noted that the bank may eventually develop agents for investment banking pitchbooks or employee surveillance.

While analysts suggest the automation could result in hundreds of millions of dollars in annual savings and potential staff reductions, Argenti characterized expectations of immediate job cuts as premature. He argued that the current goal is to inject capacity to improve the client experience and business growth, though he noted that third-party providers could face cuts as the technology evolves.