IIIBC · AI Strategy Institute · Technology-Driven Business Edition
AI Readiness Assessment for Technology-Driven Businesses
A diagnostic for growth-stage technology operators — SaaS, fintech, healthtech, legaltech, marketplaces, and AI-enabled platforms. Answer honestly across eight dimensions; you will receive a scored readiness report and have it emailed to your leadership team. AI should compound your advantage, not your risk.
How to use this assessment
Designed for CTOs, CEOs, VPs of Engineering, Chief Data/AI Officers, and board technology committees. Each question is scored 0–3: 0 = not in place / no awareness · 1 = awareness only · 2 = in progress / partial · 3 = fully in place and practiced. There are no trick questions — this is a diagnostic, not a grade. Your answers are saved automatically in this browser, so you can close this page and return to finish later.
Quick pre-check — the 5 questions every technology leader should be able to answer
If your team cannot answer these confidently, this assessment will tell you exactly why.
- What AI tools are currently in use across the organization — and who approved them?
- When an AI system produces a wrong or unexpected output, what is the escalation path?
- Has leadership defined what AI is not allowed to decide without human review?
- Can you name the one AI initiative in the next 12 months that would most change your competitive position?
- Who owns AI strategy accountability today?
IIIBC · Beyond Cognition
AI Readiness Report
Prepared for your company
Overall Readiness Profile
0/159
—
What this means
Dimension Scorecard
| Dimension | Your Score | Max | Readiness Level |
|---|---|---|---|
| Total | 0 | 159 |
Readiness by Dimension
Priority Gap Analysis
Dimensions scoring Foundational or Developing are your highest-leverage next moves.
Recommended AI Use Cases
Recommended Next Step with IIIBC
Investor & Board Conversation Guide
Framings for discussing AI strategy and ROI with your board and investors.
| Board / investor concern | Your response |
|---|---|
| "Where is the ROI on AI spend?" | "Every material initiative is mapped to a board KPI on a scorecard, with X% of AI workflows now in production — not pilots — and measured impact on cycle time, cost-to-serve, or margin." |
| "Are we exposed on data or security?" | "We have an acceptable-use policy and shadow-AI controls, an AI model inventory and risk register, and our controls map to the OWASP LLM Top 10 and NIST AI RMF." |
| "Will AI erode our gross margin?" | "We track AI cost as a % of COGS and contribution margin by segment, with caching, model routing, and usage-aligned pricing to protect margin as usage grows." |
| "What's our competitive moat in AI?" | "Our proprietary transactional, behavioral, or clinical data compounds with usage and grounds product features that generic models cannot replicate." |
| "Are we compliant as regulation lands?" | "We risk-tier every AI system against the EU AI Act and sector rules, and we treat ISO 42001-style governance as a sales advantage, not just a cost." |
| "Why not just build it all ourselves?" | "We build only where AI is a true differentiator and buy the commodity layer — buying/partnering succeeds roughly twice as often as internal builds at our stage." |
Appendix — Sources & Calibration
This assessment is calibrated against current (2024–2026) research and standards. Key references used to ground the questions and thresholds:
Maturity & adoption
- Gartner — AI Maturity Model & Roadmap and AI production-durability survey (Jun 2025): 45% of high-maturity orgs keep AI in production 3+ years vs. 20% of low-maturity.
- McKinsey QuantumBlack — The State of AI (2024; 2025): leaders, not employees, are the bottleneck; workflow redesign is the strongest predictor of EBIT impact; only ~20% measure AI with business metrics.
- MIT NANDA — The GenAI Divide: State of AI in Business 2025: 95% of generative-AI pilots show no measurable P&L return; buy/partner succeeds ~67% vs. ~33% for internal builds.
- BCG — AI Adoption Puzzle / AI at Work 2025 and the 10-20-70 operating-model rule (70% of value is people/process).
- a16z — Generative AI in the Enterprise (2024) and AI Application Spending Report (2025): bottom-up tool adoption; internal vs. customer-facing deployment rates.
Data, platform & people
- Gartner — AI-ready-data forecast: ~60% of AI projects abandoned through 2026 due to data not being AI-ready (~80% for GenAI); data-lineage adoption.
- MLOps/LLMOps maturity literature (arXiv 2025; practitioner LLMOps guides): version everything, CI/CD with eval gates, production monitoring/drift.
- IBM / agility-at-scale — CAIO role failure patterns; hub-and-spoke operating models ~36% higher ROI than decentralized.
- Workday (2026, via CIO.com) — ~40% of AI time savings lost to reworking AI output when roles are not redesigned; HBS — talent-density amplification; WEF — workforce/skills shift.
Security & reliability
- OWASP — Top 10 for LLM Applications (v2025) and Top 10 for Agentic Applications (2026): prompt injection, sensitive-information disclosure, improper output handling, excessive agency, vector/embedding weaknesses, unbounded consumption.
- IDC / IBM / Harmonic Security — shadow AI: 56% of employees use unauthorized AI tools; sensitive-data exposure on consumer tiers; ~$670K higher average breach cost with high shadow-AI.
- LLM observability/eval practice — hallucination and silent drift require quality monitoring beyond uptime; red-teaming now expected.
Regulation
- EU AI Act — four-tier risk model; staggered obligations (prohibited practices + AI literacy from Feb 2025; GPAI from Aug 2025; high-risk + transparency + GPAI enforcement from Aug 2026; product-embedded high-risk from Aug 2027); penalties up to €35M or 7% of global turnover. A 2026 "Digital Omnibus" amendment may shift high-risk dates — verify current status.
- NIST — AI Risk Management Framework 1.0 (Govern / Map / Measure / Manage) and the Generative AI Profile.
- ISO/IEC 42001:2023 — certifiable AI management system.
- US patchwork — NYC Local Law 144 (AEDT bias audits, live since 2023); California AI Transparency Act (SB 942 / AB 853); Colorado AI Act (postponed and, in 2026, repealed-and-replaced — verify the replacement law); revised federal bank model-risk guidance (2026, superseding SR 11-7).
- Sector overlays — fintech (model-risk management), healthtech (HIPAA + FDA SaMD / PCCPs), legaltech (ABA Formal Opinion 512; UPL).
Value & economics
- FinOps / CloudZero (2025) — ~84% of companies report AI infrastructure cutting gross margin by 6%+; AI-native product margins trending well below traditional SaaS.
- Drivetrain / CFO practice (2025) — track cost per workflow, AI cost as % of COGS, contribution margin by segment; power-user concentration as a margin risk; balanced AI scorecard with outcome-appropriate cadence.
Engineering transformation (IIIBC field synthesis)
- Theory of constraints applied to engineering: ~30–40% of effort is coding, 60–70% is surrounding work; a 10× coding speedup yields only 2–3× overall without process redesign — measure where time goes first.
- Spec-first / verification-first delivery: "AI executes; the engineer specifies and verifies." Specs become the new code review; build verification infrastructure before scaling AI code generation.
- Multi-agent chains (planner → implementer → reviewer) over single trusted agents; agent-ready codebases (monorepos,
AGENT.md, clean boundaries) as a prerequisite. - Engineering AI maturity & cost curve (L1–L5, ~$20 → ~$20,000 per engineer per month) and the new measurement stack (AI code share, complexity-adjusted velocity, code turnover ratio, cost per ticket/feature, DX pulse).
- New risk classes: bad code at scale, activity bloat, runaway token spend, single-vendor dependency, agent tech debt. Enterprise alignment (procurement, security, legal, finance, HR) is the gating factor; accountability for outcomes remains human.
Calibration caveats. Several 2026 regulatory items (EU Digital Omnibus, Colorado repeal-and-replace, revised US bank model-risk guidance) were in flux at publication; verify the specifics applicable to your markets. Pilot-failure statistics (95% / 85% / 88%) come from different studies with different definitions and are directionally — not precisely — comparable.