
You have completed the AI Control Assessment. You know where governance stands today, how the institution compares to peers, and which of the four strategies — Rent, Rent + Govern, Compose, or Build — applies to the institution.
Now the harder question: is the chosen strategy resilient across the multiple futures AI is creating simultaneously?

AI does not create one future. It creates multiple plausible futures — arriving faster, and in less predictable patterns, than institutions are built to manage.
Consider two cases:
The Build commitment that becomes unnecessary. An institution commits significant capital to fully-owned AI infrastructure based on current regulatory expectations. Three years later, regulatory developments make sovereign infrastructure unnecessary for that institution's profile. Capital deployed; competitive position compromised.
The Rent + Govern strategy that loses its provider. An institution builds AI Control around enhanced contractual terms with its primary AI provider. Two years later, the provider is acquired, restricted by export controls, or becomes politically untenable. The institution has no fallback.
Both institutions made defensible strategic decisions based on conditions at the time. Both decisions broke under conditions that scenario planning would have surfaced — and that resilient strategies could have anticipated.
The AI Control Assessment tells you where you are and where you need to go. Scenario Planning tells you whether the path you have chosen remains sound when the assumptions underlying it change.

A rigorous, decision-led methodology developed at the University of Oxford and applied to institutional AI strategy.
The Oxford Scenario Planning Approach (OSPA) is not forecasting. It is not prediction. It is a structured method for building a small set of plausible future operating contexts so that leadership can stress-test decisions, surface hidden assumptions, and choose strategies that remain resilient across change.
Applied to institutional AI governance, OSPA answers four questions:
1. What AI future are we implicitly planning for — without realizing it?
2. What breaks in our chosen strategy if that future does not arrive?
3. Where do we need control, resilience, and optionality built into the programme now?
4. What should we build, buy, partner on, or exit — under different futures?
These are the questions Stage 1 Assessment produces direction on. Stage 2 Scenario Planning produces resilience around.

Five structured steps from decision framing to actionable triggers.
STEP 1 — FRAME THE DECISION.Identify what leadership must decide. Strategic decisions worth scenario planning typically involve significant capital commitment, multi-year horizons, or institutional positioning that is difficult to reverse.
STEP 2 — IDENTIFY CRITICAL UNCERTAINTIES.Surface the forces shaping AI outcomes that the institution cannot control: regulatory direction, technology evolution, geopolitical developments, market structure, vendor consolidation, and others specific to the institution's context.
STEP 3 — BUILD SCENARIOS.Construct plausible operating contexts (typically 3–4) that combine critical uncertainties into coherent futures. Scenarios are not predictions of which future will arrive — they are stress-test environments for strategic decisions.
STEP 4 — TEST STRATEGY.Evaluate the institution's current strategic direction against each scenario. Identify which strategic elements hold across all scenarios (resilient), which break under specific scenarios (vulnerable), and which require redesign (fragile).
STEP 5 — TRANSLATE INTO ACTION.Produce concrete outputs: no-regret moves (actions that make sense in every scenario), strategic options (decisions that depend on which scenario emerges), trigger indicators (early signals that one scenario is materializing), and decision rights (who acts when triggers fire).

A fast, leadership-ready engagement to create 3–4 credible AI futures and translate them into immediate strategic choices.
Outputs

We pressure-test your current AI roadmap against multiple futures and redesign it for resilience.
Outputs

We use scenarios to strengthen accountability, oversight, model risk management, vendor risk, and decision rights.
Outputs

We help you avoid lock-in and build a partner ecosystem that works across futures.
Outputs

Scenarios become an operating tool—not a workshop artifact.
Outputs

A SCENARIO SET. Three to four plausible futures, constructed as coherent operating contexts. Each scenario describes regulatory developments, technological evolution, market structure, and competitive dynamics over the relevant time horizon. Scenarios are written for board-level audiences — clear, narrative-driven, and grounded in observable forces.
A VULNERABILITY MAP. The institution's current AI strategic direction tested against each scenario. Strategic elements that hold across all scenarios (resilient), elements that break under specific scenarios (vulnerable), and elements that require redesign (fragile).
A PORTFOLIO OF OPTIONS. Concrete strategic alternatives across build, buy, partner, and exit decisions — each calibrated to which scenarios make it the right choice.
NO-REGRET MOVES. Actions that make sense regardless of which scenario emerges. These are the decisions the institution can act on now without resolving scenario uncertainty.
DECISION TRIGGERS. Early-warning indicators tied to specific scenarios, with associated decision playbooks. Establishes who acts, when, and how — before the institution is forced into reactive responses.
A BOARD-QUALITY DELIVERABLE. All outputs structured as a single board-ready document supporting fiduciary oversight, regulatory examination, and ongoing executive decision-making.
Your roadmap holds across multiple AI futures, reducing surprise and rework.
Risk, accountability, and decision rights become explicit—so AI doesn’t remain a “black box.”
Investments are sequenced and optioned—avoiding overbuild, lock-in, and wasted spend.
Teams move from debate to disciplined choices, backed by shared scenarios and triggers.
You design for independence, portability, and continuity as the AI landscape shifts.
This page describes Institutional AI's Stage 2 Scenario Planning engagement model as of April 2026. Engagement durations, formats, deliverables, and outputs described on this page are illustrative of typical Institutional AI Stage 2 engagements. Actual engagements are calibrated to each institution's specific strategic, operational, and decision context, and may vary materially from descriptions on this page.
The Oxford Scenario Planning Approach (OSPA) is a methodology developed at the University of Oxford. Institutional AI applies the methodology to institutional AI strategy under appropriate professional training; references to Oxford or OSPA do not imply endorsement, partnership, or affiliation with the University of Oxford.
Scenarios developed in Stage 2 engagements are plausible future operating contexts for stress-testing strategic decisions. Scenarios are not predictions, forecasts, or guarantees of future conditions. Decisions made on the basis of scenario analysis remain the institution's responsibility.
Stage 2 outputs are intended to support institutional decision-making and do not constitute legal, regulatory, investment, tax, or fiduciary advice. Institutions should consult appropriate professional advisors before acting on scenario findings.
Information provided for informational and educational purposes only.
AI Control. For Financial Institutions.
© 2026 Institutional AI. All Rights Reserved. OLTAIX™ and The Institutional AI Stack™ are trademarks of Institutional AI. Provided for informational purposes only and does not constitute legal, regulatory, investment, or other professional advice.