加强加勒比地区的人工智能准备（英）

Strengtheningtheartificial intelligence Introduction Artificial intelligence(AI)systems, when applied well,couldhelp in addressingkey globalchallenges and advance the Sustainable Development Goals(SDGs).However, these potentialbenefits will not materialize automatically.The use of AI also carries significant risks, such asdisruption to job markets and economies, loss of cultural diversity, new AI-enableddigital While AIsystemscould support the achievement of many of the SDGs ,theycould alsoimpedethe achievement ofothers andincrease inequalities within and betweencountries (Vinuesaandothers).Many experts in the field of AI are concerned thatexisting or new harms resultingfromAI will become substantially more serious or widespreadin the near future, due to inequalities Bearing in mindthese risks andthe promises of the 2030 Agenda for Sustainable Developmentto “leave no one behind”, “reach the furthest behind first” and“reduce inequality within andamong countries”,it is clear that governments must strengthen theircapacityto govern AI.Thereal harmsassociated with AImustbe mitigated,newrisksand threatsmustbe anticipated,andits potential and opportunitiescouldbe harnessed for the benefit of people across the Caribbean governments areactivelyengaging with the globalAI policy and governancedialogues, but governance gaps still exist at the national level.This policy briefdiscusses theCaribbean’s AI readinessbyreviewingaction taken by Caribbean governments, as well asdiscussinginfrastructure,digital skillsandthe private sector. It alsosituatesAI and the digitaltransformation in the contextof the ongoing triple planetary crisis of climate change, CARIBBEAN ARTIFICIAL INTELLIGENCEREADINESS Policies, strategies and laws Caribbean governments are actively engaged with the global discussions on how to regulate AI.This isevidenced by the unanimous adoption of the UNESCORecommendation on the Ethics of Artificial Intelligence(2021), andthat several Caribbean States co-sponsored the two2024General Assembly resolutions onAI(78/265 and78/311).At the regional level,the Digital Agenda for Latin America and the Caribbean (eLAC2026)speaks tofostering innovation, emerging technologies and AI for sustainable development,whilesafeguarding human rights and the ethical use of technology.The Santiago(2023)and Cartagena(2024)declarations on AI, which were signed by a handful of Caribbean countries each,discuss key elements of AIgovernance. This includesestablishingeffective national institutionalframeworks anddevelopingecosystemsfor the development and deployment of ethical, safe and inclusive AI. At the subregional level, theCaribbeanArtificialIntelligencePolicy Roadmap,which was developed in 2021 and updated following extensive regional Atpresent, however,the development of AI strategies is still in the emerging stagesat the nationallevel.Policies are reported to be in development in a few countries, but only one country has adopted anational AI strategy as yet.The manifold efforts of Caribbeancountries is nevertheless evidence oftheir desireto establish a strategic direction for the subregion on AI.Their concerted efforts to deepeninternational andmultistakeholder collaborationto maximize the benefits and minimize the risks associated withAI are laying Laws and regulations on AI arein the very early stages of development globally, and no country orregion is yet ready to fully regulate AI. Moreover, no one currentlyunderstands all oftheinner workingsof AIenough to fully control its outputs or predict its evolution(UN HLAB,2024b), which presents a formidablechallenge to regulators worldwide.Given the uncertainties surrounding AI technology,the rapid pace ofdevelopment, themanyvaried documented harms and risks, and thefact that AI policy development is still inits early stages in most countries, AIlegislation is likely to take some time todevelop in the Caribbean. countries and territories, with some bills having been introduced(Stanford AI Index, 2024).Puerto Ricoandthe United States Virgin Islandshavealso adopted at least one AI-related law each (NSCL,2024). At the international level, theCouncil of EuropeFramework Convention on Artificial Intelligence,Human Rights,Democracy and the Rule of Law(AI Treaty), which was opened for signature in September 2024,is the only legally binding international treatyon AIcurrently in existence. However,no Caribbean country hassigned theTreaty as yet, though Caribbean countries should consider whethersuch signature might beadvantageous. Theapplication of theTreaty to the Caribbean territorieswill bedetermined by theirspecificconstitutional arrangementswith their administering power, orwill beclarified duringtheratification process, Whileit might be too early to adopt AI-specific legislation in the Caribbean at present,this is a goodtime to build the capacity of legislators andregulators related to digital technology and AI. Legislatorsin thesubregionalso have plentyof topicsto consider in