AI news ·
Siloed patient data and disconnected clinical systems limit what AI can deliver, research finds
Four enterprise debts are trapping nearly $18 trillion in recoverable value across Global 2000 firms, with healthcare and life sciences carrying a $1.2 trillion revenue impact. Only 6% of organizations have programs actively tackling these structural obstacles to scale AI.
A new report from Genpact and HFS Research identifies four enterprise debts trapping nearly $18 trillion in recoverable value within Global 2000 companies, with healthcare and life sciences carrying some of the heaviest burdens. Based on a survey of more than 2,000 senior executives across 16 industries, the research shows that for payers and providers, the biggest challenge is not whether AI can work, but whether existing operating models and workflows can enable scaling.
The four debts holding AI back
Healthcare organizations face structural obstacles rooted in decades of layered-in core systems, including electronic health records. The researchers categorize these as technology debt, data debt, process debt and talent debt. Technology debt consumes approximately 42% of developer time servicing legacy infrastructure rather than generating returns. Data debt means 42% of AI and analytics initiatives are already failing across industries because enterprises lack AI-ready data. Process debt, where inefficient or manual workflows eat up roughly 40% of employee time each week, creates a risk that ungoverned AI workflows will encode existing inefficiencies into automated systems. Talent debt, while lower in healthcare than in other industries, still leaves workforce AI-readiness estimated at just 32%.
"This is why our core conviction is that there is 'no artificial intelligence without process intelligence,'" said Balkrishan Kalra, president and CEO of Genpact. "The companies that commit to this work will not gain a few points of advantage. They will gain market share by a factor."
The courage gap
Most organizations are aware of these debts but are not acting on them. More than 50% of respondents lacked initiatives to resolve their debts and realize higher returns on AI investments. Just 6% had established and are measuring results from programs that tackle institutional challenges to implementing AI at scale. "The gap between them and the 94% is not a planning gap," the researchers said. "It is a courage gap."
Healthcare and life sciences carry the second largest revenue impact at $1.2 trillion and cost impact of $2.1 trillion, behind only manufacturing. The sector runs "the most complex multiparty workflows in the global economy, meaning process debt accumulates at every handoff," the report said.
A dual-velocity approach
Lisa Stump, Mount Sinai Health System's chief digital information officer, suggested that agentic AI might help leapfrog past some of healthcare's entrenched debts while the foundational work continues. "Can agentic AI compensate for the messy workflows … at least in the short term? Can it operate across less-than-perfect data, multiple systems and clunky workflows while we do the hard work to clean the data and streamline the work?" Stump said. "In essence, we need to act on both for near-term, albeit not perfect, wins and long-term sustainable value."
The researchers outlined five key learnings from the 6% of organizations that are actively resolving these debts. They recommend treating enterprise debt as a CEO mandate, not an IT project; operating at dual-velocity to fix foundations while scoring short-term wins; using AI to fix what AI needs to run on; not under-investing in talent; and recognizing that action beats ambition.
Why this matters for healthcare professionals
Fragmented data environments and legacy clinical-administrative systems drag down employee readiness for AI, and the stakes extend beyond lost investment value. Faulty data can lead to misdiagnoses or missed diagnoses that put patients at risk. For healthcare leaders, the report's message is direct: deploying agentic AI before redefining process workflows and building data foundations does not accelerate transformation - it automates existing inefficiencies. The path forward requires confronting structural debt with process intelligence and a mandate that reaches beyond the IT department. Professionals looking to build these capabilities can explore resources like AI for Healthcare Courses and AI for Executives & Strategy to close readiness gaps at both the operational and leadership levels.