The world’s top 2,000 public companies are sitting on nearly $18 trillion in trapped AI value, according to new research from Genpact and HFS Research. The value is not blocked by the AI models. It is blocked by the organizations attempting to implement them.
The study, which surveyed more than 2,000 enterprise executives across 16 industries and 14 functions, identifies a condition the researchers call “enterprise debt”: the accumulated backlog of data quality issues, process inefficiencies, outdated technology, and workforce readiness gaps that AI surfaces and amplifies rather than bypasses. Most companies have been managing these debts for years. AI deployment is making them structurally urgent in a way that business-as-usual tolerated.
The Four Enterprise Debts Blocking AI at Scale
Genpact and HFS Research identify four categories of enterprise debt that collectively explain why AI pilots fail to graduate into production.
Advertisement
300 × 250
Data debt is the most immediate. Only 33 percent of enterprise data is currently AI-ready, and 42 percent of AI initiatives are failing at least partly due to data quality problems. The AI model works; the data it is working with does not.
Process debt is the second category. Forty percent of employee time is lost weekly to inefficient and manual processes. AI tools can automate some of those processes, but when the underlying workflows are poorly documented or inconsistently executed, automation locks in the inefficiency rather than eliminating it.
Technology debt creates a third constraint. Enterprise core systems average 10 years old, and 42 percent of developer time is being consumed servicing existing technical debt rather than building new capability. Deploying AI on top of aged infrastructure is an integration problem at every step.
Talent debt is the fourth barrier. Only 32 percent of the enterprise workforce is currently AI-ready, and as ICIMS data published this week shows, the supply of AI-savvy talent is not keeping pace with demand. Computer programmer openings grew 35 percent year-over-year, software developer openings grew 28 percent, and database administrator openings grew 27 percent, while overall hiring grew only 1 percent. The talent the organizations need to deploy AI at scale is also the talent the market is fighting over most aggressively.
Who Is Getting It Right, and What They Are Doing
Only 6 percent of surveyed leaders qualify as “proven debt resolvers” with established, funded programs for addressing enterprise debt systematically. Eighty-five percent say their debts actively constrain AI value generation. More than 50 percent have no funded resolution plan in motion.
“AI is exposing every weakness enterprises have spent decades learning to live with,” said Phil Fersht, founder and CEO of HFS Research.
The study calculates that organizations that successfully resolve their enterprise debts can achieve 8 percent faster annual revenue growth and 16 percent annual cost reductions, which is where the $18 trillion figure originates: the aggregate of recoverable value across the Global 2000 if their debt loads were addressed.
The 6 percent of proven resolvers are distinguished primarily by having established measurement systems for their debt, funded programs to address specific debt categories, and executive accountability for resolution timelines. They treat enterprise debt as a strategic risk, not an IT maintenance backlog.
What This Means for the HR Leader
HR leaders occupy an unusual position in the enterprise debt landscape. They are a primary stakeholder in talent debt (workforce AI readiness and skills development), a major beneficiary of process debt resolution (HR administration is heavily manual), and a principal architect of the change management that determines whether debt resolution programs succeed or stall.
The workforce readiness gap in particular has a direct connection to the AI skills confidence gap already documented in enterprise settings. Organizations where employees understand AI tools, trust their own judgment in using them, and have psychological safety to experiment are the same organizations most likely to build the internal capability to resolve data and process debt over time.
For CHROs evaluating how to position HR’s contribution to AI implementation, the enterprise debt framing offers a useful structure. The question is not just “how do we train our people on AI.” It is “which specific debts is the workforce gap blocking us from resolving,” and what programs are funded to address them at the speed the business requires.
The gap between 6 percent who are resolving and 85 percent who are constrained is not primarily a technology or funding gap. It is an organizational capacity and prioritization problem. AI is already changing how talent is identified and evaluated; the next change is whether HR can also help organizations build the internal systems that let the AI they deploy actually work.
Source: PR Newswire
