Goldman Sachs has begun using AI agents for accounting and compliance across parts of its internal finance operations. The bank is deploying autonomous agents to handle structured accounting and regulatory workflows. The initiative was confirmed by Goldman Sachs Chief Information Officer Marco Argenti. The system is built using Anthropic’s Claude AI model and has been developed over approximately six months.
The AI agents are integrated directly into Goldman Sachs’ internal systems. They operate within the firm’s existing compliance and security frameworks. Human oversight remains in place for final decision-making.
How AI Agents for Accounting and Compliance Are Used
The AI agents for accounting and compliance are designed to automate repetitive, rules-based finance tasks. These include transaction reconciliation, trade accounting, and data validation. The agents also assist with compliance checks, client onboarding reviews, and regulatory reporting workflows.
The system processes large volumes of structured financial data. It compares records against predefined regulatory and accounting rules. The goal is to reduce manual work while maintaining accuracy and auditability.
Goldman Sachs is testing these agents in controlled environments before wider deployment. Each use case is reviewed to ensure compliance requirements are met.
Oversight and Risk Controls in AI Agent Deployment
Goldman Sachs maintains human supervision over all AI-driven workflows. The AI agents for accounting and compliance do not make final regulatory or financial decisions independently. Employees review outputs and intervene when required.
The bank has emphasized that the agents function as internal tools rather than external-facing systems. All operations are restricted to Goldman Sachs’ secured infrastructure. This approach is intended to limit operational and regulatory risk.
Broader Adoption of Agentic AI in Finance
The deployment reflects a broader trend of AI adoption in corporate finance. Research cited by PYMNTS Intelligence shows that approximately 45% of chief financial officers already use AI in structured finance functions. Another 52% allow AI systems to recommend financial timing or workflow adjustments.
Goldman Sachs has previously expanded internal AI assistants across its workforce. These tools support tasks such as document review, data analysis, and internal research. The accounting and compliance agents represent a more advanced application of autonomous AI within regulated financial environments.
The bank continues refining the system while monitoring regulatory expectations. The initiative marks one of the most developed uses of AI agents for accounting and compliance in global banking.