Goldman Sachs Explores AI Agents With Anthropic * Science and Technology Times

Goldman Sachs is developing AI‑powered agents with Anthropic to automate internal banking tasks.
The collaboration focuses on operational processes such as trade accounting and client onboarding.
Early results suggest the technology could significantly reduce processing time across key functions.

Goldman Sachs Expands Its AI Ambitions

Goldman Sachs is working with Anthropic to build AI agents designed to streamline a growing set of internal operations. The partnership has been underway for six months, with Anthropic engineers embedded directly inside Goldman’s teams. Their goal is to automate tasks such as trade and transaction accounting, along with client due diligence and onboarding. The bank confirmed the details following a report from CNBC.

The project uses Anthropic’s Claude model as the foundation for these autonomous agents. Goldman sees potential for the technology to accelerate workflows that traditionally require extensive manual review. Early development is still in progress, but the bank expects meaningful efficiency gains once the tools are deployed. Chief Information Officer Marco Argenti said the agents should reduce the time needed to complete core processes.

Anthropic has been expanding its enterprise offerings, including products like Claude Cowork that execute computer‑based tasks for office workers. These tools are part of a broader push to integrate AI into everyday business operations. Goldman’s adoption reflects a growing trend among financial institutions exploring AI for operational automation. The bank plans to launch its agents soon, though no specific timeline has been provided.

AI Agents Target Core Banking Workflows

The collaboration focuses on areas where automation can reduce repetitive work. Trade and transaction accounting involve large volumes of data that must be processed accurately. AI agents could help validate entries, reconcile discrepancies and flag anomalies for human review. These improvements may shorten turnaround times while maintaining compliance standards.

Client due diligence and onboarding are also key targets for automation. These processes require gathering documents, verifying identities and checking regulatory requirements. AI systems can assist by extracting information, organizing records and identifying missing elements. Goldman expects these capabilities to support employees rather than replace them, enabling staff to focus on higher‑value tasks.

Argenti noted that the bank is still refining the technology before wider rollout. Testing includes ensuring that the agents behave reliably in complex financial environments. The bank must also evaluate how the tools integrate with existing systems. These steps are essential for maintaining operational stability in a highly regulated industry.

Anthropic’s involvement suggests a focus on safety and controlled deployment. The company has positioned Claude as a model designed for responsible use in enterprise settings. This aligns with the financial sector’s need for predictable and auditable AI behavior. The partnership reflects a cautious but forward‑looking approach to automation.

A Broader Shift Toward AI in Finance

Goldman’s work with Anthropic is part of a wider movement across the financial industry. Banks are increasingly exploring AI to handle documentation, compliance checks and operational workflows. These areas are well‑suited to automation because they involve structured processes and large datasets. AI agents can help reduce bottlenecks and improve consistency.

Anthropic has been actively pursuing enterprise partnerships as demand for AI‑driven automation grows. Its tools are designed to support white‑collar tasks that require reasoning and structured output. Financial institutions represent a major opportunity due to their reliance on documentation and regulatory processes. Goldman’s adoption may encourage other firms to explore similar collaborations.

The bank has not disclosed when the new agents will be fully deployed. Argenti indicated that the launch is approaching but did not provide a firm date. The timeline will depend on testing, integration and regulatory considerations. Once operational, the agents could become a model for future AI‑enabled workflows in finance.

Goldman’s confirmation of the project underscores its commitment to modernizing internal systems. The bank has been investing in technology to improve efficiency and reduce operational risk. AI agents represent the next step in that evolution. Their success could influence how financial institutions adopt automation in the coming years.