Goldman Sachs is collaborating with AI startup Anthropic to build autonomous agents that automate a growing set of internal banking functions, CNBC reported. The bank confirmed the report, saying the effort is aimed at streamlining core operations and reducing processing time.
According to Goldman’s chief information officer, the firm has spent the past six months working alongside Anthropic engineers embedded within its teams. The agents under development target tasks such as trade and transaction accounting, client due diligence, and onboarding—areas that are typically labor-intensive and time-consuming.
Anthropic is expanding its enterprise footprint with tools like Claude Cowork, designed to execute computer-based tasks for white-collar roles. Goldman is still in early stages of deploying agents built on Anthropic’s Claude model, but expects meaningful efficiency gains once they go live. While no launch date was disclosed, the bank indicated the rollout is approaching.
The partnership underscores accelerating adoption of agentic AI in financial services, as large banks test whether autonomous systems can reliably handle regulated workflows while maintaining accuracy and compliance.
