J.P. Morgan says GenAI should accelerate quantity of work beyond 55%

J.P. Morgan estimates that generative AI could deflate IT services project costs by approximately 35% through automation of routine coding. To offset this and prevent net revenue loss, the brokerage projects a necessary acceleration of IT services workloads by over 55%. The analysis highlights a significant shift in IT work composition, with coding's share dropping from 40% to 6% post-AI, indicating a reallocation of labor towards new tasks like prompt engineering and increased code review. J.P. Morgan believes this 55% workload increase is achievable, citing historical patterns of enterprise reinvestment in new technologies and early market signals.

Analysis

J.P. Morgan's analysis indicates a significant structural shift for the IT services sector driven by generative AI, projecting a potential 35% deflation in project costs on a like-for-like basis. This is primarily attributed to a 90% cost reduction in routine coding, which comprises 40% of typical project hours. To counteract this price compression and avoid net revenue loss, the brokerage calculates that a minimum 55% increase in IT services workload is necessary. The composition of IT work is expected to transform, with coding's cost share plummeting from 40% to 6%, while new roles like prompt engineering emerge to constitute 23% of the post-AI project cost structure and code review's share doubles. Despite the deflationary pressure, J.P. Morgan maintains an optimistic outlook, believing the 55% workload growth target is attainable. This view is supported by historical precedent where cost savings from prior technology shifts, such as cloud and offshoring, were reinvested into new initiatives, and by early evidence like Genpact's reported 3% revenue lift from clients adopting new agentic solutions.

Key Decisions for Investors

• Investors should assess IT service providers on their strategic ability to stimulate new demand and retrain their workforce for higher-value tasks like prompt engineering and complex code review, as this will be critical to offsetting price deflation.
• Monitor IT service firms for key performance indicators that demonstrate a workload increase of over 55%, such as growth in project volume and complexity, which is the threshold J.P. Morgan identifies as necessary to neutralize margin pressure from AI-driven efficiencies.
• Consider overweighting positions in companies that show early evidence of successfully monetizing generative AI, such as the 3% revenue lift cited at Genpact, as this signals an ability to capture reinvestment spending and lead the sector's evolution.