Ar­ti­fi­cial In­tel­li­gence (AI) has shift­ed from be­ing an ex­per­i­men­tal HR tool to a core en­abler of work­force trans­for­ma­tion. Over the past two years, gen­er­a­tive AI (GenAI), agen­tic as­sis­tants, and ro­bust gov­er­nance frame­works have re­de­fined re­cruit­ing, learn­ing, work­force plan­ning, and em­ploy­ee ser­vices.

From chat­bots to

in­tel­li­gent agents

Ear­ly HR chat­bots were lim­it­ed to an­swer­ing FAQs and rout­ing re­quests. To­day’s AI-pow­ered agents ex­e­cute mul­ti-step work­flows across HRIS, ATS, and helpdesk sys­tems. They draft job de­scrip­tions, screen re­sumes, sched­ule in­ter­views, sum­marise man­ag­er notes, and rec­om­mend learn­ing paths.

McK­in­sey’s 2025 analy­sis high­lights these sys­tems’ abil­i­ty to plan and act au­tonomous­ly—con­firm­ing el­i­gi­bil­i­ty, check­ing bud­gets, and up­dat­ing records. HR self-ser­vice has evolved in­to “do-it-for-me” au­toma­tion.

Adop­tion and im­pact

AI adop­tion in HR is ac­cel­er­at­ing. SHRM’s 2025 re­port shows a sharp rise in or­gan­i­sa­tions us­ing AI for HR tasks, along­side a widen­ing skills gap. Glob­al sur­veys re­veal com­pa­nies re­design­ing work­flows and cre­at­ing gov­er­nance roles to con­vert GenAI pi­lots in­to mea­sur­able busi­ness im­pact.

Com­pli­ance takes cen­tre stage

Reg­u­la­tors are scru­ti­n­is­ing AI in em­ploy­ment de­ci­sions. The EU AI Act bans “un­ac­cept­able-risk” sys­tems, en­forces trans­paren­cy for gen­er­al-pur­pose AI, and im­pos­es strict rules on high-risk us­es like re­cruit­ment and work­er man­age­ment. In the US, the De­part­ment of Labour’s AI Prin­ci­ples urge em­ploy­ers to safe­guard work­er rights. New York City’s Lo­cal Law 144 man­dates bias au­dits and can­di­date no­ti­fi­ca­tions for au­to­mat­ed hir­ing tools. These frame­works de­mand fair­ness, trans­paren­cy, and ac­count­abil­i­ty in HR AI.

Caribbean con­text: Mo­men­tum and chal­lenges

The Caribbean is ad­vanc­ing AI gov­er­nance through UN­ESCO’s Caribbean AI Pol­i­cy Roadmap and the CTU’s re­gion­al AI Task Force, which pro­mote in­clu­sive dig­i­tal lit­er­a­cy and re­spon­si­ble AI adop­tion. Cari­com and part­ners are in­vest­ing in dig­i­tal trans­for­ma­tion and skills de­vel­op­ment. How­ev­er, in­fra­struc­ture gaps and un­even AI ex­po­sure—of­ten con­cen­trat­ed in ur­ban, for­mal jobs—pose chal­lenges. HR lead­ers must tar­get AI to aug­ment scarce skills, im­prove dig­i­tal flu­en­cy, and ad­dress band­width and da­ta-qual­i­ty con­straints.

Da­ta pro­tec­tion and HR AI

Caribbean ju­ris­dic­tions are tight­en­ing pri­va­cy laws:

Ja­maica: The Da­ta Pro­tec­tion Act and 2024 Reg­u­la­tions re­quire law­ful bases be­yond con­sent and bal­anc­ing tests for le­git­i­mate in­ter­ests.

Bar­ba­dos: The Da­ta Pro­tec­tion Act en­forces prin­ci­ples like pur­pose lim­i­ta­tion and da­ta min­imi­sa­tion.

Trinidad & To­ba­go: Reg­u­la­tions are be­ing fi­nalised, sig­nalling ur­gency for com­pli­ance.

HR teams must align AI ini­tia­tives with these laws, en­sur­ing law­ful da­ta use, re­ten­tion lim­its, and se­cu­ri­ty.

Five AI use cas­es for

HR lead­ers

Tal­ent ac­qui­si­tion: AI ac­cel­er­ates sourc­ing and screen­ing but risks bias if mis­used. Best prac­tice: de­fine job-rel­e­vant fea­tures, con­duct bias au­dits, and com­bine AI short­lists with hu­man judge­ment. Caribbean em­ploy­ers ben­e­fit from cross-bor­der sourc­ing but must man­age da­ta pri­va­cy across ju­ris­dic­tions.

Skills in­fer­ence & in­ter­nal mo­bil­i­ty: AI-dri­ven skills graphs match em­ploy­ees to stretch roles and learn­ing paths, boost­ing re­ten­tion. In the Caribbean, this sup­ports mo­bil­i­ty across sub­sidiaries and agen­cies—pro­vid­ed da­ta ac­cu­ra­cy and ex­plain­abil­i­ty are main­tained.

Per­son­alised learn­ing & de­vel­op­ment: AI cu­rates ca­reer-aligned learn­ing paths, clos­ing skills gaps quick­ly. Re­gion­al lead­ers should pri­ori­tise light­weight, of­fline-ca­pa­ble con­tent to over­come band­width con­straints.

Per­for­mance & pay de­ci­sions: AI can sum­marise feed­back and flag risks but falls un­der high-risk scruti­ny glob­al­ly. Doc­u­ment pur­pos­es, con­duct im­pact as­sess­ments, and en­sure hu­man over­sight.

Work­force plan­ning: GenAI com­bines in­ter­nal HR da­ta with ex­ter­nal sig­nals to mod­el staffing sce­nar­ios. Caribbean em­ploy­ers can use ECLAC fore­casts for macro in­puts while en­forc­ing pri­va­cy con­trols.

Risks HR must man­age

Bias: His­tor­i­cal da­ta re­flects in­equal­i­ty. Bias au­dits be­fore and af­ter de­ploy­ment are es­sen­tial.

Pri­va­cy: Avoid over-col­lec­tion of sen­si­tive da­ta; com­ply with re­gion­al laws.

Ac­cu­ra­cy & ex­plain­abil­i­ty: Pre­vent AI “hal­lu­ci­na­tions” by con­strain­ing mod­els to ver­i­fied sources and re­quir­ing hu­man sign-off.

Labour im­pacts: Stud­ies sug­gest 2–5 per cent of jobs in LAC could be au­to­mat­ed, while 26–38 per cent will be af­fect­ed. HR must pair au­toma­tion with reskilling to main­tain trust.

Caribbean pri­or­i­ties

for 2025–2026

Start with gov­er­nance: Adopt an AI Use Pol­i­cy aligned to UN­ESCO prin­ci­ples and NIST’s risk frame­work.

An­chor on da­ta pro­tec­tion: Map law­ful bases and com­plete DPIAs for high-risk use cas­es.

Build AI lit­er­a­cy: Train ex­ec­u­tives, HR prac­ti­tion­ers, and em­ploy­ees on AI ba­sics and rights.

Fo­cus on quick wins: Pi­lot re­cruit­ing con­tent gen­er­a­tion, skills tag­ging, and case-res­o­lu­tion agents.

Tie L&D to eq­ui­ty: Tar­get AI lit­er­a­cy and pow­er skills; mea­sure con­ver­sion from learn­ing to role changes.

Mea­sure & im­prove: Use score­cards track­ing time-to-hire, bias au­dits, mo­bil­i­ty rates, and em­ploy­ee sen­ti­ment.

Glob­al sig­nals to watch

EU AI Act: High-risk oblig­a­tions for re­cruit­ment and work­er man­age­ment are phas­ing in.

US stan­dards: NIST’s frame­works and DOL best prac­tices of­fer com­pli­ance blue­prints.

Labour mar­ket trends: LinkedIn and WEF high­light AI lit­er­a­cy as cen­tral to ca­reer mo­bil­i­ty.

A 12-month HR play­book

• In­ven­to­ry all HR AI tools and clas­si­fy risk.

• Pub­lish an AI Use Pol­i­cy and em­ploy­ee no­tice.

• Con­duct at least one ex­ter­nal bias au­dit.

• Launch a three-tier AI lit­er­a­cy pro­gramme.

• Pi­lot safe, high-val­ue use cas­es with hu­man over­sight.

• Tight­en pri­va­cy con­trols and ven­dor agree­ments.

• Mon­i­tor out­comes and share re­sults with stake­hold­ers.

Look­ing ahead

Short-term (1–3 years): AI will au­to­mate cler­i­cal tasks and aug­ment HR func­tions like re­sume screen­ing and em­ploy­ee queries. Caribbean sec­tors like tourism and ad­min­is­tra­tion will see ear­ly adop­tion.

Medi­um-term (3–7 years): AI will re­shape com­pli­ance, bank­ing, and pro­fes­sion­al ser­vices glob­al­ly, with Caribbean em­ploy­ers ac­cel­er­at­ing reskilling.

Long-term (7–15 years): Au­tonomous sys­tems, AI-dri­ven health­care, and cre­ative au­toma­tion will dom­i­nate glob­al­ly. Caribbean suc­cess hinges on dig­i­tal in­fra­struc­ture and in­clu­sive poli­cies to avoid deep­en­ing in­equal­i­ty.

The strate­gic pay­off: Done thought­ful­ly, AI en­ables HR to fo­cus on trans­for­ma­tion rather than trans­ac­tions—mov­ing tal­ent faster, teach­ing skills faster, and spot­ting risks ear­li­er. In the Caribbean, se­lec­tive am­bi­tion and strong gov­er­nance will un­lock pro­duc­tiv­i­ty while safe­guard­ing rights and in­clu­sion.

The Hu­man Re­source Man­age­ment As­so­ci­a­tion of T&T (HRMATT) is the lead­ing voice of the hu­man re­source pro­fes­sion lo­cal­ly. HRMATT Says is a col­umn meant to ad­dress is­sues and con­cerns of pro­fes­sion­als and the gen­er­al pub­lic, fo­cused on hu­man cap­i­tal de­vel­op­ment. To­day’s ar­ti­cle is writ­ten by HRMATT mem­ber and vol­un­teer on the PR Com­mit­tee, Jonathan Cum­ber­batch. Learn more about HRMATT by vis­it­ing our web­site: www.hrmatt.com. Fol­low us on Face­book, LinkedIn, In­sta­gram and Twit­ter. Con­tact us at: 687-5523 or via email: sec­re­tari­at@hrmatt.com