Marco Argenti, Goldman Sachs’ chief information officer, outlined a new phase in the bank’s AI journey where the emphasis shifts from saving money to creating new business value through artificial intelligence.
- Goldman moves from cost-cutting to AI-driven growth
- Developers become managers of AI agent teams
- Focus on secure, accountable AI deployment
What happened
Goldman Sachs chief information officer Marco Argenti spoke at the Wave by Vento event in Turin, detailing the bank’s evolving AI strategy. Since joining Goldman in 2019 from Amazon Web Services, Argenti has overseen a rapid integration of AI across the organization’s workflows. He described three distinct phases of AI adoption: an initial phase focused on early adopters among developers, a second phase aimed at reengineering processes for efficiency, and the current third phase focused on driving growth by leveraging AI to generate revenue and business innovation.
According to Argenti, the initial AI use cases primarily served developers experimenting with new tools, while the subsequent phase rethought operational workflows to enable straight-through processing with minimal human intervention. The present stage empowers teams to complete projects faster, fund AI initiatives through tangible returns, and shift developers’ roles toward managing AI agents that can autonomously create sub-agents. This enables developers to act more like entrepreneurial managers, prioritizing outcomes and resource coordination.
Why it matters
The move from cost-saving to growth marks a significant maturation in AI deployment at a major financial institution. By emphasizing projects that deliver measurable business value—such as shortening project timelines and enabling self-sustaining AI initiatives—Goldman Sachs is demonstrating AI’s transition from experimental use to strategic asset. This evolution reflects broader trends in the industry where AI is not merely a tool for efficiency but a driver of competitive advantage.
Moreover, Argenti’s description of developers managing AI agent hierarchies points to a redefinition of technical roles within the bank, where human expertise is augmented by AI autonomy. This expanded dynamic requires new skills in outcome definition and oversight rather than direct coding, fundamentally changing how innovation happens at scale in finance.
What to watch next
Goldman Sachs’ continued investment in secure AI adoption frameworks is notable. Argenti emphasized ‘zero trust’ and ‘defense in depth’ strategies that govern where and how AI agents operate, ensuring accuracy and safety are built into workflows to mitigate risk. The bank’s approach recognizes the intrinsic variability of AI outputs and builds safeguards accordingly, including AI models monitoring each other to catch errors.
Industry observers should also monitor how Goldman balances open and proprietary AI models, leveraging open-weight models for flexibility and cost-efficiency while deploying frontier models for complex, high-stakes tasks. This dual approach may become a blueprint for financial institutions seeking both innovation and control. Finally, the organizational changes Argenti highlighted suggest ongoing shifts in talent management and project funding that could influence broader financial sector AI adoption trends.