AI-Wiki

ai-maturity-measurement-comparison

5 min read

Thread: How do different sources measure AI maturity, and where do they agree or diverge?

The question

Multiple authoritative sources are publishing adoption and maturity numbers for enterprise AI, using different instruments, samples, and definitions. Sometimes they appear to converge (“most orgs aren’t really mature”); sometimes they appear to diverge (78% adoption vs. 7% Stage 4). Is there a coherent meta-story across these sources, or are they measuring fundamentally different things?

What we know so far

From AI Index 2025 (citing McKinsey survey, n=2,854)

  • 78% of orgs use AI in at least one business function (vs. 55% in 2023).
  • 71% of orgs use generative AI in at least one function (vs. 33% in 2023).
  • Only 1% of C-suite executives describe their GenAI rollouts as “mature.”
  • Most reported financial impact: cost savings <10%, revenue gains <5%.
  • Methodology: McKinsey Global AI Survey instrument; “use” = at least one business function; “mature” = self-reported by C-suite.

From MIT CISR (Future Ready Survey 2022, n=721, plus 2024 interviews at 9 enterprises)

  • Four-stage AI maturity distribution: Stage 1 (Experiment) 28%, Stage 2 (Pilots) 34%, Stage 3 (Ways of working) 31%, Stage 4 (Future-ready) 7%.
  • Stages 1–2 below industry-average financial performance; Stages 3–4 above.
  • Maturity score = equally-weighted combination of three measures: AI effectiveness for (i) operations, (ii) customer experience, (iii) ecosystem support.
  • Greatest financial impact in moving Stage 2 → Stage 3.

Apparent convergence

QuestionAI Index (McKinsey)MIT CISRConvergence
Are most orgs “using AI”?78% (any function)100% (Stages 1–4 are all using AI in some sense)Both say yes — high adoption breadth
Are most orgs mature?1% mature (C-suite self-report on GenAI)7% Stage 4Both say no — maturity is rare
Is there a clear value gradient?Cost savings <10%, revenue <5%Stages 1–2 below avg, Stages 3–4 aboveBoth say yes — but CISR locates the inflection (Stage 2→3)

Apparent divergence (or methodological)

  • Definition of “mature.” AI Index uses C-suite self-report on GenAI specifically. CISR uses a triangulated AI-effectiveness score across operations, CX, and ecosystem support — and mature is Stage 4, the top 7%. These are not the same metric.
  • Sample. McKinsey n=2,854 global, 2024. CISR Future Ready n=721, 2022 — 2 years older. The 2022 baseline predates the GenAI explosion; current CISR distributions might look quite different.
  • Adoption granularity. AI Index: binary (used in ≥1 function or not). CISR: 4 stages with explicit progression criteria. The CISR view is richer but the AI Index view is more comparable across years.

Two more instruments (added 2026-04-28 batch)

From Cisco (Cisco’s 2025 readiness survey)

A third measurement type — readiness and urgency, distinct from adoption breadth or maturity stage:

  • 13% of companies globally are ready to leverage AI to its full potential.
  • 68% say their infrastructure is at best moderately ready.
  • 98% feel increased urgency over the past year.
  • 85% give themselves <18 months to deploy a strategy or face negative business effects.
  • 50% have already dedicated 10–30% of IT budget to AI.

These numbers are complementary, not competing with AI Index/CISR. They describe a gap analysis between what orgs are doing (high adoption, low maturity) and what they think they need to do (massive urgency, short deadlines, high IT spend). Caveat: Cisco’s survey instrument; Cisco-favorable framing on “infrastructure readiness.”

From Werner-Le-Brun (HBR Nov–Dec 2025)

A broader org-change baseline, not AI-specific:

  • Just 12% of transformation efforts show sustainable performance gains, even after three years.
  • Despite trillions in investment over the past 20 years.

This contextualizes the AI maturity numbers. If only 12% of general transformations succeed, then CISR’s 7% Stage 4 + AI Index’s 1% mature is actually consistent with — possibly even worse than — the broader transformation-success baseline. AI maturity may be even harder than general transformation, given the speed of capability change and the depth of org-redesign required.

Updated cross-walk

SourceWhat it measures”Successful” %Sample / instrument
AI Index 2025 (McKinsey)GenAI rollout maturity (C-suite self-report)1% maturen=2,854, 2024, McKinsey instrument
MIT CISR (2022 baseline)Stage 4 “AI future-ready” (triangulated AI-effectiveness score)7% Stage 4n=721, 2022, CISR Future Ready Survey
Cisco (2025)“Ready to leverage AI to full potential”13% readyCisco’s instrument; details thin
Werner-Le-Brun (HBR 2025)Transformation efforts → sustainable performance gains12%Aggregated from cited research

The four “success rate” numbers (1% / 7% / 12% / 13%) span a single order of magnitude despite radically different instruments — which is itself useful information: whatever you call success, it’s rare.

What’s missing / candidate sources to find

  • Updated CISR Future Ready Survey wave (post-2022, ideally 2024). If it exists in the wiki later, ingest it to see if Stage distributions shifted.
  • Other organizational maturity frameworks — e.g., Gartner’s AI maturity model, Deloitte’s State of AI in the Enterprise, McKinsey’s own “AI high-performer” framing. As of this thread opening, none have been ingested.
  • A direct MIT Sloan/CISR ↔ AI Index methodology cross-walk — if anyone has published one. Not yet found.
  • MITTRI_Cisco (in raw/reports/, not yet ingested) is sponsored MIT Tech Review Insights research, likely to introduce a third framing. Will surface in the next ingest.
  • HBR Anand-Wu Gen AI Playbook (in raw/articles/, not yet ingested). Authors are at NYU Stern / HBS — different academic lineage, may have a different framing.

How this thread should resolve

When at least 3 sources are ingested with explicit adoption/maturity definitions, this thread should resolve into a syntheses page that:

  1. Lists each source’s instrument, sample, and definition side-by-side.
  2. Maps each source’s “mature” / “high-performer” definition to the others.
  3. Notes where the numerical claims agree vs. disagree.
  4. Identifies what the practical implication is for someone trying to assess where their own org stands.

Until then, this thread stays open as a reminder to always include the instrument when citing an adoption/maturity number — “78% of orgs use AI” without saying “per McKinsey’s at-least-one-function definition” loses important meaning.

Graph View

Backlinks