← AI Native Landscape 🕐 6 min read
AI Native Landscape

MIT CISR Enterprise AI Maturity: The Stage 2→3 Transition Is Still the Highest-Value Move

Brandon Sneider · April 2026

The August 2025 update (Report XXV-8) provides the most granular financial performance data to date, measured as deviation from industry-average growth and profit:

See also (wiki): ai-maturity-models, workflow-redesign, it-operating-models, training-architecture, ai-roadmap-execution

Executive Summary

  • MIT CISR’s Stephanie Woerner presented updated Enterprise AI Maturity findings at the AsiaPac International Executive Forum on March 26, 2026, extending the framework first published in December 2024 (n=721) and updated in August 2025 (n=152, 19 executive interviews).
  • The four-stage model continues to show a hard financial break between stages 2 and 3. Organizations in stages 3–4 post above-industry-average financial performance; those in stages 1–2 run below average. The spread between the bottom and top stages is roughly 30 pp of revenue growth and 20 pp of net margin.
  • As of the 2025 survey wave, 64% of enterprises had reached stages 3 or 4 — up from 38% in 2022. The migration is real. But 36% remain stuck in pilot mode, and the data says that is costing them measurably.
  • The August 2025 update introduced the “4S” framework for the Stage 2→3 transition: Strategy, Systems, Synchronization, Stewardship. None of the four is a technology purchase. All four are organizational architecture decisions.
  • Two new case studies — Guardian Life Insurance ($14.5B revenue) and Italgas Group (€1.78B revenue, Europe’s largest gas distributor) — illustrate what Stage 3 execution looks like in practice, including specific cycle-time improvements and commercialization of internal AI capabilities.

The Financial Break: Updated Numbers

The August 2025 update (Report XXV-8) provides the most granular financial performance data to date, measured as deviation from industry-average growth and profit:

Stage Name 2022 % 2025 % Growth vs. Industry Profit vs. Industry
1: Experiment & Prepare Exploration, education 28% 13% −26.5 pp −15.1 pp
2: Build Pilots Business cases, proven POCs 34% 23% −6.8 pp −1.4 pp
3: AI Ways of Working Scaled platforms, redesigned work 31% 46% +4.7 pp +0.8 pp
4: AI Future-Ready Continuous innovation, new revenue 7% 18% +13.9 pp +9.9 pp

Total spread: 40.4 pp of growth and 25.0 pp of profit between Stage 1 and Stage 4.

The Stage 2→3 transition remains the highest-leverage move: it flips financial performance from below-average to above-average. But the data also shows that Stage 3 is not yet where the large returns live — Stage 4 organizations (18% of the sample) post nearly 3x the profit premium of Stage 3.

Methodology note: The 2025 survey (n=152) is materially smaller than the 2022 base (n=721). Financial performance is measured relative to industry averages, which controls for sector differences. The AI Effectiveness scale combines operational improvement, customer experience enhancement, and ecosystem support into a 0–100% composite. Stage cutoffs: 1 (0–49%), 2 (50–74%), 3 (75–99%), 4 (100%).

The 4S Framework: What the Stage 2→3 Transition Requires

MIT CISR identifies four organizational capabilities — none of which are technology purchases — that gate the transition from proven pilots to scaled value:

Strategy: Aligning AI investments with measurable, strategic goals. Stage 2 organizations run pilots that demonstrate value. Stage 3 organizations fund only work that connects to a quantified business outcome. Guardian Life assigned accountability to its Data and AI team for selection and prioritization, then built a three-phase value-tracking framework: hypothesis → business case → scaling plan.

Systems: Architecting modular, interoperable platforms and data ecosystems. Stage 2 organizations build one-off solutions. Stage 3 organizations build reusable infrastructure. Italgas invested in a platform approach since 2017: IoT (7M+ smart meters), data platform (300TB+), 23 deployed AI models, and self-service BI for business users. Components are shared between the Digital Factory innovation hub and operating business units.

Synchronization: Creating AI-ready people, roles, and teams while redesigning work around AI. This is the hardest of the four. Italgas ran 30,000+ training hours through its internal academy in 2024 and maps 95% of employees against a digital leadership model. Guardian Life pulled personnel from regular roles for concentrated AI use-case work in the short term while building long-term training and rotation programs.

Stewardship: Embedding compliant, human-centered, transparent AI practices by design — not layering them on after deployment. Guardian Life codified risk, legal, and compliance barriers with specific mitigations and runs a dual-track review system: formal Architecture Review Board for high-stakes deployments, fast-track board (technical risk, data privacy, cybersecurity) for lower-risk work. Italgas created a dedicated AI Director role reporting jointly to the CHRO (retitled Chief People, Innovation, and Transformation Officer) and CIO.

Case Studies: What Stage 3 Looks Like in Practice

Guardian Life Insurance Company of America

  • Profile: $14.5B revenue, $2.4B operating income (2024). CEO Andrew McMahon since 2020.
  • Stage 3 marker: RFP and quoting process compressed from 5–7 days to 24 hours via AI automation. Moving from pilot to enterprise scale in 2026.
  • Architecture: Microservices and API-first design for reusability. Cross-functional teams organized by product with end-to-end accountability. Legacy mainframe core systems upgraded; enterprise data consolidated.
  • Governance: Two-track architectural review — formal board for high-stakes, fast-track for routine. Evaluation criteria: data privacy, customer risk, operational risk, regulatory risk, cyber risk.

Italgas Group

  • Profile: Europe’s largest natural gas distributor. 12.9M customers across Italy and Greece. €1.78B revenue, €506.6M net profit (2024).
  • Stage 3–4 marker: WorkOnSite predictive AI for construction site management delivers 40% faster project completion and 80% fewer inspections. DANA generative AI system for network control. Both deployed, not piloting.
  • Commercialization (Stage 4): WorkOnSite commercialized through subsidiary Bludigit — €3M revenue at 50%+ margin in 2024. This is the Stage 4 signal: AI capabilities become a revenue source, not just a cost lever.
  • Workforce: 1,000+ employees involved in innovation initiatives since 2018. Italgas Academy delivered 30,000+ training hours in 2024. 95% of employees mapped to digital leadership model.

Source credibility: HIGH. MIT CISR is a non-vendor academic research center. The case studies include specific financial and operational metrics from named executives. Limitation: both case studies are large enterprises ($1B+ revenue), not mid-market. The framework design is scale-agnostic, but the case evidence skews large.

Key Data Points

Metric Value Source Date
Stage 1→4 growth spread 40.4 pp MIT CISR Report XXV-8 (n=152) Aug 2025
Stage 1→4 profit spread 25.0 pp MIT CISR Report XXV-8 (n=152) Aug 2025
% of enterprises at stages 3–4 64% MIT CISR (n=152) 2025
% of enterprises at stages 3–4 38% MIT CISR (n=721) 2022
Guardian Life RFP cycle time reduction 5–7 days → 24 hours MIT CISR case study 2025
Italgas project completion acceleration 40% MIT CISR case study 2025
Italgas inspection reduction 80% MIT CISR case study 2025
Italgas AI training hours (2024) 30,000+ MIT CISR case study 2025
Italgas AI commercialization revenue €3M at 50%+ margin MIT CISR / Bludigit 2024

What This Means for Your Organization

The MIT CISR data delivers one clear message: the value of AI is not in the technology — it is in the organizational architecture that scales it. Every company in Stage 2 has proven AI works. The 36% still stuck there are not lacking proof of concept. They are lacking the four organizational capabilities (Strategy, Systems, Synchronization, Stewardship) that convert proven pilots into enterprise-wide financial performance.

The practical question for a mid-market CEO or CIO is: which of the 4S capabilities is your binding constraint? Most companies intuitively reach for Systems (buy a platform, build a data lake). The CISR evidence suggests Synchronization — redesigning roles and teams around AI, not just distributing tools — is the one that stalls the most organizations. Italgas invested 30,000 training hours and mapped 95% of its workforce to a digital competency model. Guardian Life pulled people from their regular roles to concentrate on AI use-case work. Neither of those is a technology purchase. Both are prerequisites for Stage 3.

If diagnosing your organization’s position in this framework — and identifying which of the four capabilities is gating your progression — would be a useful exercise, I would welcome the conversation: brandon@brandonsneider.com.

Sources

  1. Woerner, S.L. “Scaling Enterprise AI Maturity for Bottom-Line Impact.” Session presentation, MIT CISR AsiaPac International Executive Forum, March 26, 2026. Member-restricted. https://cisr.mit.edu/publication/2026_0326_ScalingEnterpriseAIMaturity_Woerner

  2. Woerner, S.L., Sebastian, I.M., Weill, P., Káganer, E. “Grow Enterprise AI Maturity for Bottom-Line Impact.” MIT CISR Research Briefing XXV-8, August 21, 2025. n=152 enterprises, 19 executive interviews. Credibility: HIGH — independent academic research, relative financial performance methodology. https://cisr.mit.edu/publication/2025_0801_EnterpriseAIMaturityUpdate_WoernerSebastianWeillKaganer

  3. Weill, P., Woerner, S.L., Sebastian, I.M. “Building Enterprise AI Maturity.” MIT CISR Research Briefing, December 19, 2024. n=721 enterprises (2022 survey). https://cisr.mit.edu/publication/2024_1219_BuildingEnterpriseAIMaturity_WeillWoernerSebastian

Brandon Sneider | brandon@brandonsneider.com April 2026