Healthcare Faces $3.3 Trillion in Enterprise Debt Blocking AI Scale

Healthcare organizations are sitting on more than $3 trillion in trapped value caused by enterprise debt that prevents artificial intelligence from delivering on its promise, according to new research examining how structural challenges block AI adoption at scale.

A joint study by Genpact and HFS Research surveyed over 2,000 senior executives across 16 industries and identified nearly $18 trillion in recoverable enterprise value across Global 2000 companies. Healthcare and life sciences ranked second only to manufacturing, carrying a $1.2 trillion revenue impact and $2.1 trillion cost impact from accumulated enterprise debt.

Four debts trap AI value

The research identifies four categories of enterprise debt that specifically hamper healthcare's AI readiness:

Technology debt consumes roughly 42% of developer time servicing legacy infrastructure rather than building new capabilities. For health systems layered with decades of electronic health record implementations and clinical systems, this debt is structural rather than temporary.

Data debt represents the gap between available data and AI-ready data. Across industries, 42% of AI and analytics initiatives are already failing due to data quality issues. Healthcare's fragmented data environments and disconnected clinical-administrative systems make this problem particularly acute.

Talent debt reflects workforce readiness for AI, estimated at just 32% across industries. While healthcare scores better than some sectors on raw talent availability, siloed systems and complex workflows drag down employee preparedness.

Process debt accounts for approximately 40% of employee time lost to inefficient or manual processes each week. The researchers warn that deploying agentic AI before addressing process workflows risks "encoding existing inefficiencies into automated systems and running them at speed."

Why it matters

Healthcare runs some of the most complex multiparty workflows in the global economy, meaning process debt accumulates at every handoff between payers, providers, labs, and other stakeholders. This structural complexity means AI readiness isn't primarily a technology question but an operational one. Organizations that attempt to scale AI without addressing underlying debts may accelerate problems rather than solve them, with potential consequences for patient safety when faulty data leads to misdiagnoses or missed diagnoses.

A dual-velocity approach

Only 6% of surveyed organizations have established programs to resolve these institutional challenges and are measuring results. The gap between this small cohort and the remaining 94% is "not a planning gap" but "a courage gap," researchers said.

Lisa Stump, chief digital information officer at Mount Sinai Health System, suggested agentic AI might help organizations leapfrog some debts while simultaneously doing the foundational work. "Can agentic AI compensate for the messy workflows at least in the short term," she asked, operating across imperfect data and clunky workflows while teams clean data and streamline processes.

The research offers five principles for organizations willing to confront enterprise debt: treat it as a CEO mandate rather than an IT project, operate at dual velocity by fixing foundations while scoring short-term wins, use AI to fix what AI needs to run on, invest adequately in talent, and prioritize action over ambition.

"You cannot out-innovate broken foundations," said Balkrishan Kalra, president and CEO of Genpact. "Understanding exactly where these debts live and how to resolve them requires context-rich process intelligence."

The findings were first reported by Healthcare IT News based on the Genpact and HFS Research study released this week.