ISO and AI management-system consultancy division of LETA International (FZE)☎ +971 58 596 8708WhatsApp
LETAADVISORYDiscuss AI governance ↗

FREE · NO EMAIL REQUIRED

AI Governance Readiness Assessment

Answer all 24 checkpoints. Choose “Yes” only when the control is implemented and you can identify credible evidence.

01Leadership & accountability
L1Each important AI use case has an approved business purpose.

Look for: Approval, business case or use-case record

L2A named manager is accountable for each AI system or use case.

Look for: Owner recorded in the AI inventory

L3Responsibilities for users, reviewers, IT, security, privacy and risk are clear.

Look for: Role matrix or documented responsibilities

L4The organisation defines approved, restricted and prohibited uses of AI.

Look for: AI policy, acceptable-use rules or equivalent

02Inventory & risk
R1The organisation inventories AI tools, embedded features and autonomous agents.

Look for: Current AI system and use-case register

R2AI risks and affected stakeholders are assessed before live use.

Look for: Risk or impact assessment

R3Higher-impact uses receive stronger approval, testing and oversight.

Look for: Risk classification and control criteria

R4Risk assessments are revisited after material changes or incidents.

Look for: Review trigger and revision records

03Data & security
D1Rules control confidential, customer and personal information entered into AI systems.

Look for: Data-handling rules and technical restrictions

D2AI agents and integrations receive only the access and permissions they need.

Look for: Least-privilege access review

D3Supplier terms covering data use, retention, training and location are reviewed.

Look for: Vendor assessment and approved terms

D4AI-related security and privacy incidents have an escalation route.

Look for: Incident process, contacts and response records

04Human oversight & operations
H1A competent person reviews consequential AI outputs before action.

Look for: Approval point in the operating process

H2Users know how to validate outputs and recognise important limitations.

Look for: Instructions, training and validation checks

H3AI systems and significant changes are tested before production use.

Look for: Test plan, acceptance criteria and results

H4The organisation can stop, override or safely fall back when AI fails.

Look for: Stop conditions, manual fallback or rollback plan

05People & transparency
P1Employees receive AI literacy and role-specific competence support.

Look for: Training plan and competence records

P2People are told about material AI use when transparency is relevant.

Look for: Notices, disclosures or user information

P3Affected people can question, challenge or complain about important outcomes.

Look for: Appeal, complaint or human-review route

P4Important AI inputs, outputs, approvals and actions are traceable.

Look for: Logs, records and retention controls

06Monitoring & improvement
M1Performance, accuracy, safety, bias and security are monitored where relevant.

Look for: Measures, thresholds and review records

M2Failures, near misses and complaints lead to corrective action.

Look for: Incident, root-cause and action records

M3AI suppliers and connected services are reviewed throughout their use.

Look for: Supplier monitoring and change review

M4Management periodically reviews AI risks, performance and improvement priorities.

Look for: Internal review, audit or management-review record

24 checkpoints remaining.Results are calculated in your browser. No answers are sent to LETA.

USE THE SCORE WELL

A percentage starts the management conversation. Evidence completes it.

The checklist draws on practical themes in AI governance: leadership, risk, data, human oversight, people and continual improvement. Review the lowest-scoring categories first and validate each “Yes” against evidence.

Read the current AI governance update →