Track export demand alongside adoption
Monitor global-services renewals, client concentration and paid work. Faster delivery cannot compensate for contracts that no longer exist.
← Caribbean economic researchJamaica: jobs, AI
Caribbean AI Association research
Jamaica: jobs, AI
and export demand
Explore the work people do, the services Jamaica sells and the conditions that could change both.
Bank teller
Wider adoption with slower worker adjustment.
Industry employment, July 202636,100Observed industry total; not this occupation’s headcount
Role labour-demand index, 203098.2Unchanged work and demand = 100; assumed inputs
Realised time capacity6.0 per 100 workersFTE-equivalent time; not positions eliminated
Routine records, reconciliations and service messages can change; cash handling, identity exceptions and vulnerable-customer support still require people.
Recommended action. Track transaction errors, unresolved exceptions, branch access and customer outcomes. Do not treat digital channels or ATMs as effects of generative AI alone.
Role assumptions and a workload calculator
Automation-eligible time25%
Augmentation-eligible time30%
Demand elasticity0.7
Role headcount in JamaicaUnavailable
Time shares and demand responses are planning assumptions, not measured exposure. The O*NET occupation and task source describes US work. Local time weights and an occupation-by-industry headcount are unavailable.
The assumed direct demand loss for this role in the selected case is 0.0%. Labour demand is indexed as 100 × (1 − direct demand loss) × productivity^(elasticity − 1), assuming full productivity-related cost pass-through. Indirect credit, supplier and income effects are not estimated.
Realised capacity = max(0, 70% × adoption × [automation share + augmentation share × (1 − 1/speed)] − 5% × adoption). The calculator updates this formula only. It does not change the stored scenario comparison or the national economic model.
THE WORK INSIDE THE OCCUPATIONBank teller:
Bank teller:
task changes
Each occupation combines physical work, information, judgement and responsibility.
Balance currency, coin, and checks in cash drawers at ends of shifts and calculate daily transactions, using computers, calculators, or adding machines.O*NET task 2549
Receive checks and cash for deposit, verify amounts, and check accuracy of deposit slips.O*NET task 2551
Monitor bank vaults to ensure cash balances are correct.O*NET task 2563
Count currency, coins, and checks received, by hand or using currency-counting machine, to prepare them for deposit or shipment to branch banks or the Federal Reserve Bank.O*NET task 2554
Examine checks for endorsements and to verify other information, such as dates, bank names, identification of the persons receiving payments, and the legality of the documents.O*NET task 2552
Resolve problems or discrepancies concerning customers' accounts.O*NET task 2560
Enter customers' transactions into computers to record transactions and issue computer-generated receipts.O*NET task 2553
Human-ledAssistedDelegation candidateNew work
Selected task examples, with provisional rule-based labels. Displayed task counts do not determine the assumed time shares. Professional approval and physical delivery are not delegated by this illustration.
NATIONAL SCENARIOSAdoption, adjustment
Adoption, adjustment
and export demand
Core economic model
Broad adoption
Output level gap+3.6%
Unemployment3.76%
Cognitive group wage gap-2.0%
Other group wage gap+2.9%
Core outcomes use a matched no-AI comparison. GDP gaps are differences in output levels, not annual growth. Wage groups are model categories, not forecasts for the selected occupation.
Scenario assumptions and comparisons
Core adoption by 2030 is 30%, autonomy is 75%, and the new-task parameter is 0.25. The assumed cognitive employment allocation is 38.4%. Adoption is a share of affected task instances, not all jobs.
Demand cases impose output weights of 3% for outsourced services and 8% for direct tourism. These are stress assumptions, not observed Jamaican GDP shares. There is no current input-output multiplier. The music shock affects the role calculation only; no music GDP share is invented. The full report specifies the equations, paired comparisons, fixed labour force and omitted feedbacks.
| Scenario | Type | GDP gap | Unemployment |
|---|---|---|---|
| Gradual adoption | Core economic model | +0.71% | 3.40% |
| Broad adoption | Core economic model | +3.61% | 3.76% |
| Rapid automation | Core economic model | +13.50% | 8.07% |
| Worker-led adoption | Core economic model | +2.56% | 3.60% |
| Digital services upgrade | Core economic model | +4.94% | 3.81% |
| Outsourcing retreat | Demand stress test | +2.06% | Not estimated |
| Nearshore outsourcing collapse | Demand stress test | +0.50% | Not estimated |
| Tourism demand shock | Demand stress test | +1.54% | Not estimated |
| Combined export shock | Demand stress test | -0.79% | Not estimated |
THE EMPLOYMENT BASELINE
Where Jamaicans work
STATIN reported 1,419,200 people employed in July 2026, with unemployment at 3.4% and informal employment at 56.2%. Employment shares describe workforce scale; they do not measure an industry’s GDP contribution or AI exposure. Source ↗
| Industry | People employed | Share |
|---|---|---|
| Wholesale and Retail Trade; Repair of Motor Vehicles and Motorcycles | 268,700 | 18.9% |
| Agriculture, Forestry and Fishing | 182,200 | 12.8% |
| Construction | 152,200 | 10.7% |
| Real Estate and Other Business Services | 146,800 | 10.3% |
| Education, Human Health and Social Work Activities | 127,300 | 9.0% |
| Accommodation and Food Service Activities | 125,500 | 8.8% |
| Arts, Entertainment, Recreation and Other Services | 117,100 | 8.3% |
| Manufacturing | 87,100 | 6.1% |
| Public Administration and Defence; Compulsory Social Security | 73,800 | 5.2% |
| Transportation and storage | 70,700 | 5.0% |
| Financial and Insurance Activities | 36,100 | 2.5% |
| Information and Communication | 12,500 | 0.9% |
| Electricity, Gas and Water Supply | 10,400 | 0.7% |
| Combined suppressed / unallocated residual | 8,800 | 0.6% |
| Total | 1,419,200 | 100.0% |
The combined 8,800-person residual preserves suppressed categories, unallocated employment and rounding. It is not a mining count. Occupation headcounts cannot be recovered from these industry totals.
ASSUMPTIONS AND UNCERTAINTY
How assumptions change the outlook
These are conditional scenarios. Adoption, task times and movement between jobs depend on assumptions; they have not been measured locally for this study. Industry demand stresses use assumed value-added weights and do not estimate national unemployment.
RECOMMENDATIONS
Priorities for the transition
Test workflows in finance and professional services
Measure accuracy, verification costs, fees and working time. Separate routine processing from professional judgement and approval.
Link tourism workflows to visitor demand
Review reservations, guest service and staffing together. Do not attribute changes caused by demand or recovery to AI alone.
Measure musicians’ income by channel
Separate live performance, recordings, publishing and licensing. Track consent, ownership and attribution as well as production costs.
Evaluate worker transitions
Include informal workers and measure placement, earnings and retention. Training completion alone does not show that people found better work.
SOURCES
Data and research references
Caribbean AI Association, version 2.0. Sources retrieved 20 October 2026. The full report records each source’s use and access scope.
- S01Korinek Jones Sacher Cotter and McCrory Economic Scenarios for Transformative AI ↗
Working Paper 2026-02; Sections 2–3 and Appendix A; task-based economic framework
- S02Anthropic Scenarios for our Economic Future ↗
Public task illustration and scenario explorer
- S03STATIN Labour Force Data Explorer ↗
Main Indicators; EMP by Industry; EMP by Occupation; EMP by Formality; INFOR_FOR EMP by Industry; EMP by ISCED. July 2026, history to January 2024
- S04STATIN Labour Force Tables downloadable workbook ↗
Historical cross-check; download ends July 2025, not current baseline
- S05STATIN National Accounts methodology ↗
2008 SNA and JIC 2016; disaggregated historical IO table available at cost
- S06STATIN GDP by expenditure and income at current prices 2015 to 2023 ↗
2023 GDP compensation taxes subsidies mixed income; JMD million; published via STATIN Annual GDP page
- S07STATIN Input Output table 2007 ↗
Historical structure only; not used as current 2026 multipliers
- S08O NET Resource Center Database 31.0 ↗
Task statements and occupation CSV; 21 representative US occupations; database CC BY 4.0
- S09ILO Generative AI and Jobs A Refined Global Index of Occupational Exposure ↗
Gmyrek et al.; DOI 10.54394/HETP0387; exposure versus displacement; need local microdata
- S10ILO ISCO 08 structure and definitions ↗
Task-based occupations; major groups and need for detailed crosswalk
- S11World Bank Digital Economy Country Diagnostic Jamaica ↗
Executive summary and contents; infrastructure skills firms and trust
- S12IMF Jamaica RFI approval following Hurricane Melissa ↗
16 January 2026; recovery and external financing context, not AI impact
- S13MIIC Jamaica global services and AI ↗
6 March 2026; more than 50000 global digital services workers reported; policy view is not causal evidence
- S14Alan Lujan Econ-ARK economic scenarios implementation ↗
MIT-licensed implementation of the task-based economic framework
- S17Acemoglu The Simple Macroeconomics of AI ↗
DOI 10.3386/w32487; aggregation of task savings
- S18Acemoglu and Restrepo Automation and New Tasks ↗
DOI 10.3386/w25684; displacement and reinstatement
- S19Brynjolfsson Li and Raymond Generative AI at Work ↗
DOI 10.3386/w31161; setting-specific workplace productivity evidence; no effect size transferred
- S20UNESCO Re|Shaping Policies for Creativity (2026) release ↗
Creative income, licensing and synthetic content competition; global projections are not transferred to Jamaica
- S21Bank of Jamaica Balance of Payments data portal ↗
External-sector data acquisition route; no BPO-only or current sector value-added share obtained
O*NET® 31.0 Database, US Department of Labor / Employment and Training Administration, used under CC BY 4.0. Task classifications, weights and new-work examples are CAIA adaptations. No USDOL/ETA or Anthropic endorsement is implied. Econ-ARK model attribution and its MIT licence are retained in the research package.