---
id: "claim-ai-roi-failure"
type: "claim"
source_timestamps: ["¶1", "¶2"]
tags: ["ai-roi", "technology-adoption"]
related: ["concept-capability-mirage", "entity-gartner", "prereq-ai-workflow-understanding"]
confidence: "high"
testable: true
speakers: ["Paola Cecchi-Dimeglio"]
sources: ["reskilling"]
sourceVaultSlug: "hbr-seg-reskilling"
originDay: 10
articleStem: "hbr-edu-33-new-tools-workforce-training"
sourceUrl: "https://hbr.org/2025/12/the-new-tools-that-can-improve-workforce-training"
sourceTitle: "The New Tools That Can Improve Workforce Training"
---
# AI Investments Will Miss Returns Due to Human Utilization Gaps

## Claim: AI Investments Will Miss Returns Due to Human Utilization Gaps

**Confidence (as asserted): high · Testable: yes**

Companies are investing **$1.5 trillion in AI initiatives this year**, projected to reach **$2 trillion by 2026**, yet most of this spending will **fail to meet expected returns** ([[entity-gartner-d33|Gartner]]). The author's key move: the bottleneck is **not the technology** but the organization's failure to train people to *use* it. Employees frequently revert to legacy tools (like Excel) months after AI implementation because traditional training fails to impart practical understanding — the [[concept-capability-mirage|capability mirage]] in action.

Understanding this claim assumes familiarity with enterprise AI integration — see [[prereq-ai-workflow-understanding]].

> **External validation & caveat:** The **directional claim is well supported** — Gartner repeatedly reports that a large share of AI projects fail to deliver expected value (failure rates often cited around **80–85%** for early AI/ML initiatives), and the "AI productivity paradox" literature (Brynjolfsson and colleagues) confirms returns depend on complementary human capital and organizational change. **However:** (1) the specific **$1.5T / $2T figures** are forward-looking analyst *projections*, not hard outcome data (IDC forecasts worldwide AI spend at ~$500–900B by mid-decade depending on definitions); and (2) Gartner attributes failure to *multiple* causes — poor change management, data quality, misaligned use cases, governance gaps — **not training alone**. Focusing on upskilling risks underweighting these structural issues. See [[appraisal-xr-targeted-not-universal]].


## Related across articles
- [[claim-ai-competence-gap]]
- [[concept-capability-mirage]]
- [[claim-infrastructure-scales-adoption]]