JPMorgan tracks engineers’ AI tool use via dashboards

JPMorgan Chase & Co. is monitoring how its software engineers use artificial intelligence tools through internal dashboards, according to details described in internal tracking systems. The approach adds pressure on developers to incorporate AI into day-to-day work and to show measurable gains tied to that adoption.

The bank’s Global Technology division has about 65,000 people. Developers have been instructed to demonstrate “meaningful improvement” in both code quality and the volume of code produced by using AI, with internal tools used to measure progress toward those expectations.

Dashboards rank engineers based on their usage of AI products including GitHub Copilot and Anthropic’s Claude. The systems display adoption and usage rates, and they also surface employee-level details such as recent AI activity, office location, and reporting structures.

As of late March, nearly 70,000 employees had been provisioned for Copilot, and about 24,000 were active users. The tracking is presented internally as a way to quantify uptake and understand how widely the tools are being used across teams.

A JPMorgan spokesperson said the data is intended to evaluate the effectiveness of AI investments and to help target training, and that it is not used for performance management. However, some developers said managers have referenced usage data during performance discussions, indicating that the metrics can still influence workplace conversations even if they are not formally positioned as evaluation inputs.

The effort aligns with a broader push across large technology-focused employers to encourage AI adoption and to validate spending on these tools. Companies including Meta and Google have also tracked and promoted AI usage among employees, reflecting a wider corporate focus on productivity and measurable returns from AI-related investments.

For JPMorgan, the initiative sits at the intersection of workforce management and technology strategy, with internal measurement designed to show whether AI tools are changing output and quality in software development. At the same time, the presence of employee-level dashboards introduces uncertainty about how consistently the data will be interpreted across teams, particularly when developers report that usage metrics have already appeared in performance-related discussions.