What is AI regulation in Haiti?
AI regulation: countries and regions
As at 8 June 2026, Haiti does not appear to have a dedicated AI law, national AI strategy or specialist AI regulator in force. AI in Haiti is governed indirectly, through the Constitution, electronic communications and signature rules, public administration digitisation measures, telecom regulation and general criminal and sector law. Regional soft-law frameworks from the OAS, and more broadly Caribbean digital policy work, are more developed than Haiti's domestic AI-specific rulebook.
Reviewed by Jackie, Head of Learning & Development, Levellers · Last reviewed 8 June 2026
What this means
AI regulation in Haiti currently means the mix of law, institutions and governance practices that affect AI use, rather than a single AI Act. In other words, the key question is not "what is Haiti's AI statute?", because there does not seem to be one. The real question is which existing rules apply when an organisation uses AI in Haiti.
That matters because AI does not operate in a legal vacuum. Even without an AI-specific law, an AI system can still affect privacy, communications secrecy, administrative fairness, record-keeping, public procurement, telecom compliance and criminal liability. Haiti's framework is therefore best understood as indirect and still developing.
This is also different from a pure privacy question. Privacy and personal data are part of the picture, but AI governance in Haiti also touches public administration, digital identity, electronic transactions, telecom oversight, cybersecurity and regional policy alignment.
Why it matters
For organisations deploying or governing AI in Haiti, the main practical risk is not breaching a detailed AI code. It is misreading a thin and fragmented rulebook. Haiti's official materials show a state still building its digital governance architecture, through electronic administration, electronic signature, telecom oversight and broader public-sector modernisation. Where AI is used in customer service, document handling, public services, screening, surveillance, education, health or security, the absence of a dedicated AI statute does not remove legal exposure. It means teams must work harder to map which existing duties still apply, who the relevant institution is, and where the law is incomplete or unsettled. For buyers, founders, advisers and public bodies, that changes how diligence should be done: less box-ticking against an AI Act, more close attention to constitutional rights, administrative process, data handling, security, human review and contractual controls.
How it works
No dedicated AI statute has been identified
Official Haitian materials reviewed for this article do not show an enacted AI-specific statute, a published national AI strategy or a dedicated AI regulator. Haiti's public digital governance documents focus instead on electronic administration, digital infrastructure, electronic transactions, electronic signatures and broader state modernisation. That is the right starting point for understanding AI regulation in Haiti: AI is discussed as part of digital transformation and future-readiness, not yet as a separate legal field with its own comprehensive act.
This makes Haiti a thin-source jurisdiction for AI regulation. Readers should not confuse that with an unregulated jurisdiction. It means the governing architecture is indirect, spread across older constitutional rules, digital administration measures, telecom oversight and general law.
What currently governs AI in practice
The most durable layer is constitutional. Haiti's 1987 Constitution protects the secrecy of correspondence and other communications, and it also restricts searches of homes and papers to lawful procedures. Those protections matter wherever AI is used for surveillance, communications analysis, automated monitoring, data extraction or state decision support. They do not mention AI by name, but they still constrain intrusive digital practices.
A second layer comes from Haiti's electronic administration framework. The 2016 decree recognising the right of any administrated person to communicate with the public administration by electronic means is important because it does more than allow digital processes. It says the use of information technologies must respect privacy, the protection of personal data, honour and reputation, and it ties electronic administration to security and proportionality principles. In practical terms, if a public body in Haiti later adds AI to digital forms, workflows, portals or internal decision support, those principles already sit in the legal background.
A third layer comes from electronic transactions and electronic signatures. Haiti's public-sector modernisation documents list the 2017 framework on electronic signatures and electronic exchanges as core legal building blocks for e-governance. In 2025, CONATEL publicly presented an amended signature framework and its implementation roadmap. For AI governance, this matters because AI systems may draft, classify or route documents, but legal validity still turns on the electronic signature and authentication framework, not on whether the document was intelligently generated.
The institutions are digital and sectoral, not AI-specific
Haiti does not appear to have a single institution that acts as the national AI authority. Instead, responsibilities are split.
The Office de Management et des Ressources Humaines, working within the wider framework of the Primature and administrative reform, has been one of the visible anchors for e-governance. Its modernisation programme lists the core legal instruments for electronic administration and notes that an OMRH measure created a "Cellule de promotion de l'administration electronique". The same programme also refers to an interministerial committee on information technology. That tells you Haiti has approached digital governance as an administrative reform question.
The Ministry of Public Works, Transport and Communications, and especially CONATEL, sit closer to the telecom and infrastructure side. CONATEL's public materials show three things clearly. First, it remains the telecom regulator. Second, it now has an important role in the implementation of the electronic signature framework. Third, its current public messaging is about digital sovereignty, cybersecurity, digital inclusion and preparation for future technologies such as 5G and AI. That is relevant governance, but it is not yet the same thing as a horizontal AI rulebook.
Other ministries and agencies matter depending on the use case. Justice and police authorities matter where AI use intersects with criminal enforcement or investigation. Ministries running digitisation projects matter where AI is layered onto public-service delivery. In short, the relevant institution in Haiti usually depends on the sector and project, not on a general AI mandate.
Data protection is a real issue, but the framework remains incomplete
Haiti's official digital-administration texts speak in the language of privacy and personal data protection, which shows that the issue is recognised. But recognition is not the same as a settled, comprehensive regime.
A high-quality United Nations regional profile states that Haiti has no legal framework for data protection in place. Read alongside Haiti's own official materials, the picture is one of acknowledged need, partial digital-era rules and institutional ambition, but not a mature stand-alone data protection system of the kind seen in some other jurisdictions. Publicly accessible official materials reviewed for this article also did not reveal an independent, specialist data protection authority.
For AI deployers, that gap matters in two ways. First, it increases legal uncertainty, especially for systems that rely on sensitive, large-scale or cross-border data use. Second, it raises governance expectations even where black-letter detail is sparse. If the law is incomplete, good practice becomes more important, not less. Teams should therefore treat privacy impact analysis, access controls, retention limits, human oversight and complaint handling as baseline governance, especially in public-facing or high-impact systems.
Regional alignment is more visible than domestic AI-specific rulemaking
At the regional level, the OAS is the clearest reference point. Its Inter-American Framework on Data Governance and Artificial Intelligence, MIGDIA, is an active initiative designed to guide member states in developing data and AI governance. The framework includes regional guidelines, a study of the state of AI and data governance in the Americas, and a model policy. Haiti is listed among the beneficiary countries. For Haiti, this is soft law and policy guidance, not directly binding domestic legislation. But it matters because it offers a realistic template for what future Haitian AI governance could borrow from.
In the wider Caribbean policy space, CARICOM has also moved further than Haiti's domestic AI lawmaking. In 2024, CARICOM Heads endorsed a Regional Digital Resilience Strategy that includes training in areas such as artificial intelligence and data analytics, and plans for a regional AI centre of excellence. Again, this is not a Haitian AI statute. It is regional policy direction. But for a country with limited AI-specific domestic codification, that regional direction is likely to matter in practice, especially for capacity-building, standards language and future public policy design.
The practical model is horizontal, cautious and still unsettled
Put together, Haiti's current model is horizontal rather than AI-specific. AI is mainly governed through existing constitutional protections, administrative digitisation rules, telecom and e-signature frameworks, and whatever sectoral law applies to the context. It is also a soft-law influenced model, because regional guidance from the OAS and wider Caribbean digital policy work is more developed than domestic AI legislation.
This also means there are important boundaries. Haiti does not currently seem to have a published risk-tier AI framework, mandatory AI incident reporting rules, foundation-model duties, dedicated conformity assessment rules or a specialist AI enforcement authority. If such measures are needed, they would likely have to come through future law reform, sector regulation or adoption of regional policy models.
For operators, the main legal question today is usually not "is this AI prohibited?" It is "which existing rights, process rules or sector duties could this AI use affect, and who would challenge it if something goes wrong?"
Examples
One concrete example is the "Haiti numerique" digital acceleration project run through the MTPTC with World Bank financing. Its purpose is to expand high-speed connectivity to government services, schools and hospitals, and to build the foundations of Haiti's digital economy. If AI tools are later added to those connected services, the immediate governance frame is still the digital-project architecture, electronic administration rules, telecom oversight and constitutional safeguards, not a dedicated AI act.
A second example is CONATEL's electronic signature rollout. CONATEL publicly states that the amended signature framework, published in 2025, gives legal effect to electronic signatures, certificates and related certification responsibilities. So if a ministry, utility or private actor uses automated or AI-assisted document workflows, the legally significant step is still the authenticated signature and certification process. AI may help draft or sort, but it does not replace the legal conditions for validity.
A third example comes from Haiti's public modernisation documents. They identify scattered digitisation projects across government, but also openly describe weak coordination, weak regulation and the absence of a clear public policy anchor for digital governance. In practical terms, that means an organisation introducing AI into Haitian public administration should expect governance work to involve several institutions, uneven process maturity and a need for stronger internal controls than the law explicitly spells out.
Common misunderstandings
"Because Haiti has no AI Act, AI is unregulated." Not correct. AI still sits inside constitutional law, electronic administration, telecom oversight, e-signature rules, criminal law and sector-specific duties.
"Haiti already has a mature personal data regime that answers most AI questions." Not on the material reviewed. Privacy and personal-data principles appear in official texts, but a comprehensive modern data protection framework does not appear settled.
"CONATEL is Haiti's AI regulator." Not exactly. CONATEL is a telecom and digital-regulatory institution with roles in electronic signatures and digital infrastructure. That is adjacent to AI governance, but it is not the same as a specialist AI authority.
"OAS and CARICOM AI frameworks automatically apply as binding Haitian law." No. They are soft-law or regional policy reference points unless and until Haiti incorporates them into domestic measures or uses them in regulatory practice.
"If an AI system is hosted abroad, Haitian law does not matter." Not safely assumed. Systems used in Haiti can still engage Haitian constitutional protections, administrative duties, telecom rules, contract terms and sectoral obligations.
Risks and boundaries
The biggest boundary is simple: Haiti is not yet a source-rich AI-law jurisdiction. There is no clear domestic AI Act, no published national AI strategy identified in the official materials reviewed, and no specialist AI regulator. That limits how far anyone should go in describing "Haiti's AI regime".
A second boundary is the incompleteness of the surrounding legal framework, especially on personal data. Haiti's official texts recognise privacy and personal-data concerns, but the publicly visible framework remains patchy. For high-impact deployments, that creates uncertainty rather than freedom. It means organisations need stronger internal governance because the external rulebook is thin.
A third boundary is institutional fragmentation. Digital governance responsibilities are spread across administrative reform bodies, the Primature, sector ministries and CONATEL. In practice, this can make implementation slower and less predictable.
There are also source limitations. Haiti's official online legal publishing is fragmented, and some public pages are difficult to access or verify consistently. This article is based on the official and high-quality sources that were publicly accessible on 8 June 2026. Before any high-stakes deployment, readers should recheck the current status of data-protection reform, electronic-signature implementing measures and any later AI-specific policy announcement.
This article is an explanatory overview, not legal advice.
What to do next
Treat Haiti as a jurisdiction where AI governance must be designed from first principles, not from a neat AI statute. Start by mapping the use case: what the system does, what data it uses, whether it affects communications, whether it supports or influences important decisions, and whether it sits inside a public service or regulated workflow.
Build controls around the most durable duties. Those include privacy and communications safeguards, secure authentication, record integrity, traceability, human review for significant decisions, and a clear route for complaints or correction. If the system touches public administration, make sure it fits the electronic administration and electronic signature framework rather than bypassing it.
Do not wait for a Haitian AI Act before setting policy. Put vendor controls, access management, retention rules, incident escalation and model-governance procedures in place now. Watch regional developments as well, especially OAS and CARICOM work, because they are likely to shape the language of future Haitian policy more than an already-settled domestic AI doctrine.
Have a question or a suggestion, or want to understand how we research and review these guides? Read about our editorial standards and how to reach us.
FAQs
Does Haiti have an AI law in force?
No dedicated AI law was identified in the official materials reviewed up to 8 June 2026.
Does Haiti have a national AI regulator?
No specialist AI regulator was identified. Relevant responsibilities are split across existing digital, telecom, administrative and justice institutions.
What rules matter most today if we deploy AI in Haiti?
The Constitution, electronic administration rules, electronic transactions and signature rules, telecom regulation, and any sector-specific or criminal rules relevant to the use case.
Is CONATEL the main institution to watch?
It is one of the key institutions, especially for telecoms, digital infrastructure and electronic signatures. But it is not a general AI authority.
Is there a comprehensive data protection regime for AI use?
The public materials reviewed do not show a settled, comprehensive modern data protection framework comparable to a full stand-alone regime.
Do OAS AI guidelines bind organisations in Haiti directly?
No. They are regional guidance and soft law. They matter as reference points, but they are not the same as domestic statute.
Should foreign AI providers treat Haiti as a low-risk legal environment because the rules are thin?
No. Thin rules often mean higher uncertainty. Constitutional rights, administrative fairness, telecom duties and contractual exposure still matter.
What is the near-term direction of travel?
The visible direction is toward broader digital governance, electronic administration, cybersecurity, digital sovereignty and regional AI-policy alignment, rather than an immediate Haitian AI Act.