Goldman Sachs CIO: Why Individual AI Usage Is the Wrong Metric for Productivity

Goldman Sachs Chief Information Officer Marco Argenti has stated that monitoring the volume of AI usage by individual employees is an ineffective way to measure productivity gains. Argenti is instead focusing on the speed at which engineering teams move from the conceptualization of an idea to its actual execution.

While the investment bank possesses the technical capability to track exactly how much each of its 12,000 engineers utilizes AI, Argenti told Business Insider that focusing on those metrics does not provide helpful insights into overall performance.

To illustrate the flaw in tracking individual usage, Argenti compared the approach to monitoring a single athlete on a sports field.

“It would be like looking at only one player on the field,” Argenti said. “Fine, this player is doing more movements, but why am I not scoring more goals? Well, because they need to pass the ball.”

Marco Argenti

Argenti indicated that a more reliable metric is the rate at which a team develops a feature, which is evidenced by a rapid reduction in a productive team’s work backlog.

The use of AI tools has shifted the workflow from the creation of static presentations for ideas to the development of real-time prototypes that can be adjusted based on immediate feedback.

“There’s zero time between idea and prototype. You kind of “3D print” software,” Argenti said.

Marco Argenti

Goldman Sachs AI Infrastructure

Goldman Sachs has integrated several AI systems into its operations. In 2024, the firm launched the GS AI Platform, which utilizes large language models from providers such as Google and OpenAI.

The platform includes a security layer designed to protect the firm’s private data. The bank has deployed an internal version of ChatGPT for staff use.

The firm also utilizes a system called Legend. This tool allows employees to search the bank’s extensive internal files using natural language, removing the requirement for users to know the specific file name or location of the data they are seeking.

Industry Divergence in AI Metrics

The approach taken by Goldman Sachs differs from other major technology and professional services firms that have tied AI adoption to performance evaluations.

According to the Wall Street Journal, Google has begun evaluating some of its software engineers based on their usage of AI tools, although managers maintain discretion over whether these metrics impact evaluations.

Accenture has taken a more stringent stance. In April 2026, CEO Julie Sweet stated that AI fluency is now a requirement for promotion within the consulting firm.

Sweet warned that Accenture is exiting employees who are unable or unwilling to undergo reskilling in AI technologies.

Employee Sentiment and Adoption

Argenti noted that while employees at Goldman Sachs initially expressed skepticism or fear regarding the integration of AI, those attitudes are shifting as the tools are adopted.

“The dominant sentiment is really a sense of empowerment. People feel almost liberated. A few weeks or months ago, of course, there was a real bit of skepticism and fear, but I correlate that to people that were not really using it,” Argenti said.

Marco Argenti