Adoption outruns control
Models, vendors and agents enter the enterprise faster than the decision rights, contracts and controls that are supposed to govern them.
SARPI Premium Executive Assessment
Available in SARPI Pro & EnterpriseAre You Ready to Lead Sovereign AI?
AI adoption is accelerating. The real leadership challenge is ensuring that your organization can scale AI without losing control over data, infrastructure, vendors, decisions, accountability, security, trust and strategic direction.
60 Executive Questions • 6 Leadership Domains • 5 Maturity Levels • Approximately 15–20 Minutes
The new leadership gap
Most organisations can describe their AI ambition. Far fewer can evidence the control, accountability and alignment that ambition depends on. That distance is the leadership gap.
Accountability is assumed, not assigned
When an AI decision causes harm, most boards cannot name the accountable executive, the approving forum, or the evidence trail.
Sovereignty is treated as a technical detail
Data residency, model portability and exit capability are procurement footnotes rather than board-level strategic positions.
Leadership is not aligned
Executives frequently disagree on how mature the organisation actually is — and that disagreement is invisible until it is measured.
The 6 leadership domains
Each domain carries an approved weight in the overall score, so the result reflects leadership consequence rather than question count.
D1. Sovereign AI Strategy & Mandate
18%Whether sovereign AI ambition is an owned, funded, decision-making mandate rather than a stated intention.
Do we govern sovereign AI as a strategic commitment with real authority?
10 questions — Pro & Enterprise members
D2. Data Sovereignty & Residency
17%Control over where data resides, who may access it, and which jurisdiction governs it across the AI lifecycle.
Do we know and control the jurisdiction of every dataset our AI depends on?
10 questions — Pro & Enterprise members
D3. Compute, Infrastructure & Supply Chain
18%Access to trusted compute, resilience of the AI supply chain, and tested exit paths from foreign dependency.
Could we keep operating if a key compute or vendor dependency was withdrawn?
10 questions — Pro & Enterprise members
D4. Model Sovereignty & Intellectual Property
17%Ownership, tuning rights, portability, and IP position across the models the organization depends on.
Do we own the AI capability we are building, or are we renting it?
10 questions — Pro & Enterprise members
D5. Talent, Capability & Institutional Knowledge
15%Internal capability to build, operate, evaluate, and govern AI without irreplaceable external dependency.
Could we run and challenge our AI estate without our current vendors?
10 questions — Pro & Enterprise members
D6. Regulatory Alignment & National Trust
15%Alignment with applicable AI regulation, cross-border obligations, and the public trust position of the institution.
Would our AI position withstand regulatory and public scrutiny today?
10 questions — Pro & Enterprise members
Sovereign AI Leadership Capability Matrix™ preview
Your result is expressed as a weighted capability matrix: one maturity position per domain, the weight it carries, and its contribution to the overall sovereign leadership score.
| Domain | Weight | Illustrative score | Maturity level |
|---|---|---|---|
| D1 Strategy | 18% | 72 | Level 4 — Managed |
| D2 Data | 17% | 46 | Level 3 — Emerging |
| D3 Compute | 18% | 38 | Level 2 — Aware |
| D4 Models | 17% | 44 | Level 3 — Emerging |
| D5 Capability | 15% | 65 | Level 4 — Managed |
| D6 Trust | 15% | 58 | Level 3 — Emerging |
Illustrative figures for explanation only — not a real organisation's result.
The 5-level maturity model
One scale, five levels, 0–100. The same ladder is used for every domain and for the overall score, so progress is directly comparable over time.
Level 1 — Exposed
0–20Sovereign AI exposure is unmanaged at leadership level. Establish executive ownership, a sovereignty mandate, and a first exposure baseline before scaling any AI commitment.
Level 2 — Aware
21–40Leadership recognises sovereignty risk but acts reactively. Convert awareness into a funded mandate, named accountability, and decision rights that bind procurement and delivery.
Level 3 — Emerging
41–60Sovereignty practices exist in pockets and depend on individuals. Standardise the operating model, close the weakest domain, and make sovereignty a gating criterion rather than a review comment.
Level 4 — Managed
61–80Sovereign AI leadership is deliberate and measured. Focus on exit-tested independence, assured evidence, and board-level reporting that survives external scrutiny.
Level 5 — Sovereign AI Leader
81–100Leadership demonstrates durable sovereign control of strategy, data, compute, models, capability, and trust. Sustain the position through periodic exit rehearsal and independent assurance.
Answer scale
- Never — 0
- Rarely — 25
- Sometimes — 50
- Often — 75
- Always — 100
Sovereign AI Capability SOWT™ preview
Every domain is classified by deterministic rules — no interpretation, no drift. Strength at 75 and above, Opportunity from 60, Weakness from 40, Threat below that.
Strength
- No domain in this example.
Opportunity
- D1 Strategy
- D5 Capability
Weakness
- D2 Data
- D4 Models
- D6 Trust
Threat
- D3 Compute
Domains D2, D3, D4 are treated as critical: falling below 50 raises a Critical Threat flag and pulls remediation into the immediate horizon — without altering your score or maturity level.
Example result
Overall sovereign leadership score
54
Level 3 of 5 — Emerging
Sovereignty practices exist in pockets and depend on individuals. Standardise the operating model, close the weakest domain, and make sovereignty a gating criterion rather than a review comment.
What leadership sees first
The weighted overall score, the maturity level, the strongest and weakest domain, and the spread between them as an imbalance signal.
Where the risk concentrates
Critical question alerts for data visibility, vendor and model exit, accountability, oversight, cyber resilience, shadow AI, incident response and override capability.
What to do next
A sequenced action plan across Now (0–30 days), Next (30–90 days) and Later (90–180 days), each item carrying a suggested executive owner.
Pro vs Enterprise
SARPI Pro
For the individual executive who needs a defensible read on their own leadership position.
- Complete all 60 questions with autosave
- Server-scored, version-stamped result
- Sovereign AI Leadership Capability Matrix™ across all 6 domains
- Sovereign AI Capability SOWT™ with a 0–180 day action plan
- Critical question alerts and suggested owners
- Own results and reports, privately held
SARPI Enterprise
CohortEverything in Pro, plus the ability to measure the whole leadership team and report to the board.
- Multi-respondent leadership campaigns by org unit
- Leadership Alignment Score and contested domains
- Anonymous cohort heatmap with small-cohort suppression
- Cohort benchmarks and privacy modes
- Optional Not Applicable responses, centrally configured
- Board-ready PDF and slide exports
Enterprise leadership intelligence
When an executive team answers the same 60 questions, disagreement becomes measurable. The cohort roll-up shows not just how mature you are, but how much your leaders actually agree.
Leadership Alignment Score
A 0–100 measure derived from weighted disagreement across the six domains.
Contested domains
Where leaders disagree materially — usually the fastest route to unblocking decisions.
Anonymous heatmap
Distribution across respondents without exposing any individual's answers.
Board reporting
Aggregate-only outputs for sponsors, with cohort suppression below the minimum size.
How it works
- 1
Choose your mode
Pro members complete the assessment individually. Enterprise customers run it as a leadership cohort across the executive team and business units.
- 2
Answer 60 executive questions
Ten questions per capability domain on a single five-point frequency scale. Progress autosaves, so the assessment can be completed in sittings.
- 3
Scored on our servers
Every score is computed server-side against methodology v2.0.0 using weighted domains, so results are reproducible and version-stamped.
- 4
Act on the board pack
Receive the Capability Matrix, SOWT classification, critical alerts and a sequenced 0–180 day action plan with named owners.
Why sovereign AI leadership matters
- Regulators increasingly ask who approved an AI decision, on what evidence, and under whose authority.
- Vendor and model concentration is now a strategic dependency, not a purchasing preference.
- Cyber resilience and operational override capability determine whether AI can be safely scaled.
- Boards are accountable for AI outcomes long before AI programmes are mature.
- Value only compounds when sovereignty, governance and capability advance together.
Inside the SARPI ecosystem
This assessment is part of the Enterprise Assessment Center™ — it shares the same scoring engine, governance, roles and reporting as every other SARPI assessment.
SARPI™ Assessment Center
The parent module that owns definitions, sessions, scoring and aggregation for every SARPI assessment.
UseCasesX™
Turn leadership priorities into a governed, prioritised AI use-case portfolio.
Risk-adjusted ROI
Model the value case behind each sovereign AI commitment before it is funded.
Methodology & governance
How scales, weights, versioning and publication controls are governed across SARPI.
Want an expert read on your result?
Our advisory team works with boards and executive teams to translate a sovereign AI leadership result into funded decisions: mandate, decision rights, vendor position, control baseline and a sequenced programme.
Frequently asked questions
Who should complete this assessment?
Board members, CEOs, CIOs, CDOs, CISOs, chief risk and transformation leaders — anyone who holds decision rights over how AI is adopted, governed and funded.
How long does it take?
Approximately 15–20 minutes for the full 60 questions. Progress autosaves, so it can be completed across multiple sittings.
Is my score calculated in the browser?
No. Scoring is server-authoritative and version-stamped, so the same responses always produce the same reproducible, auditable result.
Can we assess our whole executive team?
Yes, on Enterprise. Multiple leaders complete the same assessment and the cohort roll-up produces a Leadership Alignment Score, an anonymous heatmap and contested-domain reporting.
Do individual answers stay private in a cohort?
Yes. Sponsors see aggregate results only, and cohorts below the minimum size are suppressed so no individual response can be inferred.
Can we see the questions before subscribing?
The six domains, weights, answer scale and maturity levels are public. The 60 questions are available to Pro and Enterprise members to protect assessment integrity.
Find out whether your leadership is ready
60 executive questions. 6 leadership domains. 5 maturity levels. A board-ready result in approximately 15–20 minutes.