Research · Updated quarterly
Caribbean AI Risk Index
A country-by-country view of where AI-related risk is concentrating, where institutional protection is improving, and what changed from one quarter to the next.
What it measures
Risk and readiness are not the same thing
A country may have strong AI capacity and still carry high exposure. CARI separates exposure from fragility before combining them into a regional risk score.
Jobs & labour
Exposure of economic activity and occupations to current AI capabilities.
Cyber & AI agents
Cyber posture, digital attack surface, ransomware exposure and AI-enabled fraud channels.
Information integrity
Exposure to synthetic media, manipulation and fragile information environments.
Governance gap
AI readiness, data protection, cybercrime law, AI strategy and public AI literacy.
Economic resilience
How well each economy can absorb AI-related disruption and concentration shocks.
Quarterly record
See what changed
Published quarters are preserved as immutable snapshots. The index now shows current score, previous-quarter score, rank movement and a separate policy-movement log.
CARI launched on 30 July 2026. Earlier comparisons will only be published when they can be reconstructed from period-correct source data and will be clearly labelled as backcasts.
Open method
Built to be challenged
Every score is published with its method, confidence coverage, sensitivity range and source logic. Policy announcements are not treated as safeguards until they meet the rules in the methodology.
View quarterly history →Researchers, national statistical offices, regulators and CERTs can submit stronger national data for review.
Submit a correction or dataset