Goldman Sachs is adopting a measured approach to the rollout of artificial intelligence, ensuring the technology is integrated safely and effectively across its operations. Currently, half of the firm’s 46,000 employees have access to AI tools, but widespread deployment remains gradual and deliberate.
Chief Information Officer Marco Argenti emphasizes that AI adoption requires careful change management, particularly as the firm experiments with agentic AI. While this form of AI could automate tasks such as compliance checks and transaction processing, Goldman Sachs is still assessing the necessary controls to prevent hallucinations and errors.
The firm’s engineering teams were the first to gain access to AI coding assistants like GitHub Copilot and Gemini Code Assist. Goldman also held internal AI competitions to encourage innovative uses of the technology. A broader rollout followed with GS AI Assistant, an in-house generative AI tool now available to 10,000 employees—including bankers, traders, and asset managers—and expected to reach nearly all employees by late 2025. The tool assists with document summarization, email drafting, data analysis, and personalized content creation.
Goldman’s cautious approach aligns with recent research from Vlerick Business School, which challenges the idea that AI can fully replace human decision-making in corporate budgeting. The study found that while AI excels in tactical budgeting, human oversight is crucial in strategic financial planning to ensure alignment with long-term business goals.
Argenti sees data quality, AI technology, and human expertise as the three pillars of a successful AI strategy. He stresses that AI should enhance human capabilities rather than replace them: “In the hands of the best people, AI will deliver the best results.”