The Numbers

You’re Running Capital Allocation on Half the Calculation

60-80%

of AI-related workforce decisions default to “replace” posture, made without coordination.

$660-725B

hyperscaler AI capex projected for 2026 (CreditSights, 2026)

128 weeks

current lead time for power transformers needed for AI data centers (Sightline Climate, 2026)

100x–10,000x

ratio of readiness diagnostic cost to documented failure recovery cost

The Exposure
The Replace Posture Is a Leveraged Bet on a Leased Production Layer
60-80% of AI-related workforce decisions in typical large enterprises are in “replace” posture — AI substituting for humans. These decisions are being made by individual function leaders, without coordination, without sensitivity analysis, and without board-level disclosure.

The upside: modest cost discipline, 1-3% EBIT improvement.
The downside: catastrophic vulnerability on infrastructure you don’t own, capability you don’t control, and pricing you can’t set.
The Framework
Both Halves of the Calculation
The Capital Cycle Trap
AI is following the classic five-phase capital cycle: Boom → Surge → Collapse → Rebound → Leveling. We are in late Phase 1 /early Phase 2 — maximum confidence, maximum danger. Phase 2 is when the visible cost of moving slowly is highest, and the cost of moving too fast is not yet visible. The augment premium costs 8-15% of Phase 1-2 savings but is structurally cheaper in Phase 3-4 by margins that historically exceed 5-20x the Phase 1-2 difference.
The CFO’s Diagnostic Questions
By function, which posture — replace vs. augment — have we adopted? What is the cost to restore eliminated capability, and on what timeline? How deeply committed are we to the current AI provider — what’s our actual switching cost? What is our explicit Phase 3scenario plan? If AI costs were 3× in 24 months, what would our operating model look like? If these answers are uncomfortable, you’ve just received the most valuable governance signal available.
What We Deliver

The Asymmetric-Risk Framework Your Board Needs

1
AI Posture Assessment

4-6 week evaluation of replace-vs-augment posture by function. Make the implicit decision explicit before it becomes irreversible.

2
Switching Cost Quantification

The real number — operational lock-in, prompt libraries, fine-tuning, team fluency — not the contract exit fee your procurement team quoted.

3
Phase 3 Scenario Modeling

Board-ready quantification of three plausible shock scenarios: 50% cost increase, capability divergence, andregulatory/geopolitical disruption.

4
Asymmetric Risk Dashboard

Make the invisible cost visible before it becomes irreversible. Phase 1-2 savings alongside Phase 3 exposure in the same units.

5
100x–10,000x Prevention Ratio

Low six-figure diagnostic vs. $100M–$7B documented failure cost. The cheapest insurance policy in the enterprise.

6
Governance Insurance

Not a consulting engagement — a fiduciary obligation. Boards that do not ask the readiness questions are accepting the certainty that they will not see failure coming.

FAQS

Frequently Asked Questions

We’ve already modeled our AI cost savings. Whatare we missing?
Isn’t the “replace vs. augment” debate just about headcount?
What’s the restoration cost if we need to reverse course?
How quickly can you deliver this assessment?
Next Step
What’s the Phase 3 Exposure on Your Current AI Investments?

You can see the Phase 1-2 savings. You cannot see the Phase 3 exposure. The asymmetric-risk framework that captures both halves is available — and implementable in 90 days. The firms that build it are the firms that survive the cycle. Every day without it, the cost of reversing your implicit decisions grows.