Sources & Research Basis

More than 80% of AI projects fail to reach meaningful production

RAND Corporation researchers conducted structured interviews with 65 experienced data scientists and engineers to investigate why AI projects fail. Their August 2024 report found that more than 80% of AI projects fail to reach meaningful production deployment — approximately twice the failure rate of IT projects that do not involve AI components.

The RAND study is notable for its methodology: rather than surveying organizational executives about their AI programs, it gathered evidence from the practitioners who build and deploy these systems. The five root causes identified — led by problem definition failure, followed by data availability failure — reflect patterns observed across real implementations rather than self-reported organizational assessments.

74% of companies have yet to unlock tangible value from AI

Boston Consulting Group surveyed 1,000 CxOs and senior executives across more than 20 sectors, spanning 59 countries in Asia, Europe, and North America, and published findings in October 2024. The report, Where's the Value in AI?, found that only 26% of companies have developed the necessary capabilities to move beyond proofs of concept and generate tangible value. Three-quarters have not.

BCG's analysis identified the key factors separating organizations that scale AI from those that do not. These factors are predominantly organizational rather than technical: change management, workflow optimization, governance, and data quality and management. Model quality ranked lower than organizational conditions as a differentiating factor.

95% of enterprise generative AI pilots deliver no measurable P&L impact

MIT's Project NANDA published findings in July 2025 based on a multi-method research design: a systematic review of more than 300 publicly disclosed AI initiatives, structured interviews with representatives from 52 organizations, and 153 survey responses from senior leaders collected at four major industry conferences. The research period covered January through June 2025.

The report found that despite an estimated $30–40 billion in enterprise investment in generative AI, 95% of organizations are seeing no measurable P&L impact. Only 5% of integrated AI pilots are extracting significant value. The authors describe these findings as a directionally accurate snapshot of enterprise AI outcomes as of mid-2025.

The authors note these findings are "a directionally accurate snapshot rather than a definitive market analysis."

Problem definition failure is the leading cause of AI project failure

RAND's 2024 analysis identified five root causes of AI project failure, ranked by frequency across the 65 practitioner interviews. The leading cause is that industry stakeholders misunderstand or miscommunicate what problem needs to be solved using AI. Projects frequently fail not because the technology is inadequate, but because the problem the technology was applied to was poorly defined, poorly scoped, or not actually suited to an AI solution.

This finding was reinforced by a secondary pattern: organizations focused on deploying the latest AI technology rather than solving a specific, well-defined business problem were disproportionately represented among failed initiatives.

Data readiness is the second leading cause

The second root cause identified in the RAND study is that organizations lack the data necessary to train an effective AI model. This encompasses data quality, data availability, and data governance — the conditions that determine whether an AI system has the inputs it needs to produce reliable outputs in a production environment.

BCG's 2024 survey of 1,000 executives independently confirmed this pattern, identifying data quality and management as the critical technology capability distinguishing organizations that scale AI from those that do not.

AI ambitions are outpacing data readiness at the C-suite level

IBM's Institute for Business Value surveyed 1,700 Chief Data Officers worldwide, publishing findings in November 2025. The study found that while 81% of CDOs prioritize investments that accelerate AI capabilities, many report that their data is still not ready to unlock AI's full potential. The authors describe this as a widening gap between AI ambition and data readiness — a gap that exists at the highest levels of organizational data leadership.

At least 30% of generative AI projects will be abandoned after proof of concept

Gartner issued a prediction in July 2024, based on a survey of 822 business leaders conducted between September and November 2023, that at least 30% of generative AI projects would be abandoned after proof of concept by end of 2025. The stated causes: poor data quality, inadequate risk controls, escalating costs, and unclear business value.

The explicit naming of inadequate risk controls alongside data quality as abandonment causes is significant. It indicates that governance failures — not only technical failures — are ending AI initiatives at the proof-of-concept stage.

Over 40% of agentic AI projects are projected to be canceled by 2027

In June 2025, Gartner issued a separate prediction for the emerging agentic AI generation: more than 40% of agentic AI projects will be canceled by end of 2027, again due to escalating costs, unclear business value, or inadequate risk controls. This prediction is based on a January 2025 Gartner poll of 3,412 webinar attendees on agentic AI investment.

The recurrence of inadequate risk controls as a named cancellation cause — now appearing in the agentic AI generation, not only earlier AI waves — indicates a structural pattern rather than an isolated issue.

Generative AI's economic potential is measured in trillions — most of it unrealized

McKinsey Global Institute modeled the economic potential of generative AI across 63 use cases in 16 business functions across 9 industries, publishing findings in June 2023. Their estimate: generative AI could add $2.6 trillion to $4.4 trillion annually to the global economy through enterprise use cases. For context, the combined GDP of the United Kingdom and Canada is approximately $4 trillion.

Set against BCG's 2024 finding that 74% of companies have yet to unlock tangible value from AI, the gap between potential and current realization is substantial. The question is not whether AI creates value — the research indicates it does, for organizations that have successfully scaled it. The question is what prevents the majority of organizations from reaching that stage.