AI ObservatoryCosta Rica
ESEN

Analysis

Costa Rica fell 17 points behind Chile in ILIA 2025

Regional benchmarking, structural capability gaps and state of the regulatory framework. Verified data from official sources and multilateral reports.

State & Algorithm

Biweekly series of analyses on AI in the Costa Rican state. Each issue unpacks a piece of the inventory: what works, what is stalled, and the open questions.

  1. 01Comparative Analysis · May 2026Costa Rica has three AI bills waiting. None was summoned.Technical comparative analysis of Costa Rica's three active AI regulation bills in the Legislative Assembly: procedural status, incompatible institutional architectures, documented MICITT opposition, and international comparison.Mario Pérez EdwardsRead article →
  2. 02State & Algorithm · No. 01AI in Costa Rica's Public Sector: 21 projects, 7 institutions, and the questions no one has answered yetAnalysis of the first systematic inventory of AI projects in Costa Rica's public sector: what exists, what returns value, what is stalled, and what lacks coordination.Mario Pérez EdwardsRead article →

Regional benchmark

ILIA score, investment and executing body for the five Latin American countries in the 2025 ranking.

  1. 01

    Chile

    70.56
    AI investment
    USD 32M
    Executing body
    CENIA (National AI Center), operational
    Key milestone
    Latam-GPT, regional 70B-parameter model built for USD 550K. UC Strategies — Latam-GPT launch coverage (Chile, USD 550K).
  2. 02

    Brazil

    67.39
    AI investment
    USD 4,000M (PBIA)
    Executing body
    Brazilian AI Plan (PBIA), led by MCTI
  3. 03

    Uruguay

    62.32
    AI investment
    Agesic with operating budget since 2007
    Executing body
    Agesic
    Key milestone
    National AI Strategy 2024-2030 approved with 40+ institutions; Ceibal program trained 80,000 students in AI. In December 2025, Decree 276/025 established the regulatory sandbox framework for AI and data spaces (supervised controlled testing environments). In April 2026, Parliament and Ceibal launched an AI training cycle for legislators and public officials. IMPO (Uruguay's Official Gazette) — Decree 276/025 establishing the governance framework for AI regulatory sandboxes (Dec. 2025). Uruguay XXI — official report on 80,000 students trained in AI by Ceibal. Sociedad Uruguaya — Parliament + Ceibal AI training cycle for legislators (Apr. 2026).
  4. 04

    Colombia

    55.84
    AI investment
    National AI Policy approved (CONPES)
    Executing body
    MinTIC + DNP
    Key milestone
    AI CONPES approved and active digital transformation agenda. MinTIC — Colombia's Ministry of Information and Communications Technologies (official site).
  5. 05

    Costa Rica

    53.83
    AI investment
    No explicit AI budget in national budget
    Executing body
    MICITT (National AI Center of Excellence pending)
    Key milestone
    ENIA 2024-2027 published (first national AI policy in Central America). Hacienda recovered ₡8,000M with AI in 2025. MICITT — Official launch of Costa Rica's National AI Strategy (ENIA).

Seven capabilities Costa Rica has not yet built

Compared with Estonia (1.3M people) and Singapore (5.6M people), two global benchmarks in digital government. Each gap cites verifiable evidence.

01

National interoperability platform (X-Road style)

Reference: Estonia

X-tee/X-Road in Estonia connects around 1,500 institutions and companies (about 600 public bodies), interoperates ~1,600 information systems and recorded close to 600 million queries last year.

Status in Costa Rica

The Costa Rican public corpus reviewed through August 22, 2026 does not document an equivalent national interoperability layer. The reviewed sources show institutional integrations but do not substantiate a shared layer of that scope.

Why it matters

Without this layer, scaling pilots like EDUS or Hacienda Digital to other institutions means building each connection from scratch. It is the largest structural bottleneck preventing AI in government from moving past isolated cases.

↗ Open
02

Formal AI governance framework

Reference: Singapore

In 2024 Singapore published a Model AI Governance Framework for GenAI with 9 dimensions (accountability, data, trusted development, incident reporting, value alignment, model evaluation, transparency, security and AI for public good).

Status in Costa Rica

Costa Rica has seven bills specifically or directly related to AI in progress (23.771, 23.885, 23.919, 24.484, 24.875, 25.171 and 25.379), four with committee reports and none passed. In the reviewed corpus, the only documented operational framework is the Judicial Branch GenAI Guidelines, which are sectoral rather than national.

Why it matters

Without a formal framework, each institution decides on its own how to evaluate risk, transparency and accountability. The uncertainty slows adoption and makes external audit difficult.

↗ Open
03

Unified national citizen chatbot

Reference: Estonia

Bürokratt is Estonia’s national conversational assistant, deployed across selected agencies and running 24/7 with voice, chat, SMS, email and sign language. Recognized by UNESCO/IRCAI as one of the Top 100 global AI projects in 2024.

Status in Costa Rica

The reviewed corpus does not document a national citizen chatbot. Hacienda's Virtual Assistant is the only documented operational case in the central government and is limited to its own domain.

Why it matters

A unified assistant reduces channel fragmentation and lowers the citizen’s entry barrier to the digital State. It is one of the highest visible-impact investments per dollar.

↗ Open
04

Internal AI assistant for public servants

Reference: Singapore

Pair, GovTech Singapore’s internal AI assistant, has 11,000+ users and 4,500+ weekly active users across the public administration.

Status in Costa Rica

The reviewed corpus does not document an equivalent government initiative. It also does not establish whether public servants use commercial generative AI tools or what institutional controls would govern such use.

Why it matters

Without an institutional tool, risk shifts to individual public servants: sensitive state data on external platforms, with no traceability. An internal tool also captures measurable productivity.

↗ Open
05

Mandatory AI testing before deployment

Reference: Singapore

GovTech Singapore runs Litmus (stress-testing for GenAI before deployment) and Sentinel (real-time content moderation) as a mandatory layer over government AI systems.

Status in Costa Rica

The reviewed corpus does not document a national pre-deployment testing framework. Available public evidence also does not support a complete comparison of the protocols applied by each institution.

Why it matters

It is the difference between an isolated AI failure and a public-trust crisis. A standard framework also lowers adoption cost for smaller institutions unable to build their own.

↗ Open
06

Numerical AI talent target with budget

Reference: Singapore

Singapore’s National AI Strategy 2.0 sets a target of tripling the AI talent pool to 15,000 people, with S$20M+ allocated to SG Digital scholarships and international internships.

Status in Costa Rica

The reviewed ENIA 2024-2027 text does not identify a numerical AI talent target with an associated budget. The catalog records LINC's more than 4,500 participants and the Ministry of Education's 90,000 teacher scholarships as capacity or digitalization programs, not as a specific national AI talent target with its own budget.

Why it matters

Without a target and budget, there is no way to evaluate progress or hold anyone accountable. Any conversation about the talent gap stays purely qualitative.

↗ Open
07

Cross-border government data agreements

Reference: Estonia

Since 2018 Estonia and Finland operate the first cross-border interoperability federation based on X-Road, with bidirectional government data flows.

Status in Costa Rica

The reviewed corpus does not document a government data federation between Costa Rica and other countries in the region. Sources reviewed on integrations with Central America and the OECD also do not substantiate operational data flows between administrations.

Why it matters

For services that cross borders (migration, trade, regional health), the lack of formal agreements forces manual processes. It is a gap that prevents Costa Rica from leveraging the regional leadership it already holds in AI dialogue.

↗ Open

7 bills, zero laws passed

The inventory covers 4 bills whose primary subject is AI and 3 directly related bills, filed between 2023 and 2026. 4 have committee reports and 3 remain in committee. Costa Rica still has no formal AI regulatory framework.

  1. 01

    File 23.771

    Law to Regulate Artificial Intelligence in Costa Rica

    General framework for the development, use and application of AI in Costa Rica. The current substitute text (September 2024) designates MICITT as the governing authority, with technical support from SUTEL, responsible for registration, oversight and auditing; it sets four risk levels and requires chatbots to disclose that the interaction is with an AI system and deepfakes to disclose that the content was artificially generated or altered. It has a majority affirmative committee report from the Science, Technology and Education Committee and has been on the Plenary agenda since October 2024, without a vote. Filed by Vanessa Castro Mora (PUSC). The bill expires on May 30, 2027 if it is not approved before then.

    Committee report issued
  2. 02

    File 23.885

    Law to Regulate New Forms of Propaganda

    Broader electoral reform intended to regulate new forms of political advertising. The bill has a committee report and reached the Plenary. Floor amendments filed against the text included rules on fake profiles, synthetic content, artificial intelligence and deepfakes; the observatory classifies it as directly related, not as a general AI law. The amendments document legislative proposals and are not presented here as enacted text.

    Committee report issued
  3. 03

    File 23.919

    Law for the Responsible Promotion of Artificial Intelligence in Costa Rica

    Framework for the responsible promotion of AI with an emphasis on fundamental rights. The current substitute text (March 2025) creates the Inter-Institutional Commission for the Development of AI (CIDIA), attached to MICITT, and an ethical, technical and scientific advisory committee; it defines the high-risk systems category and penalizes the use of AI to spread disinformation through linguistic content and to create video, image or voice files made to appear authentic in order to deceive. It has a majority affirmative committee report from the Human Rights Committee and has been on the Plenary agenda since April 2025, without a vote. Filed by Óscar Izquierdo Sandí (PLN). The bill expires on September 6, 2027 if it is not approved before then.

    Committee report issued
  4. 04

    File 24.484

    Law for the Implementation of Artificial Intelligence Systems

    Risk-based general framework. The current substitute text (October 2025) assigns MICITT the governing role and coordination with seven other institutions, and reaches foreign individuals and legal entities whose systems have effects in the country. It is the only one of the three general bills that expressly defines generative artificial intelligence as a general functional category, in addition to deep fakes. It has a unanimous affirmative committee report from the Science, Technology and Education Committee (October 2025) and has been on the Plenary agenda since February 2026, without a vote. Filed by Johana Obando Bonilla and three other legislators. The bill expires on August 6, 2028 if it is not approved before then.

    Committee report issued
  5. 05

    File 24.875

    Law to Regulate the Use of Artificial Intelligence in Electoral Processes

    Establishes disclosure duties for AI-generated content and proposes prohibitions on electoral deepfakes, automated fake accounts and microtargeting strategies designed to undermine electoral integrity. It provides for fines from 10 to 50 base salaries. At the official April 7, 2026 cutoff it remained in committee, with no substitute text or committee report; later motions were recorded, but no official evidence of a status change was found. Filed by Vanessa De Paul Castro Mora (PUSC).

    In committee
  6. 06

    File 25.171

    Law for the Protection of Personal Traits and Artistic Creations in Digital Environments

    Directly related bill covering the protection of artistic creations, personal identity and electoral transparency in digital environments. The reviewed text includes duties to identify content created or altered through AI and rules against deepfakes and impersonation. On April 21, 2026, the Supreme Electoral Tribunal objected to the electoral language then under consultation on technical and constitutional grounds. The bill remains before the Science, Technology and Education Committee.

    In committee
  7. 07

    File 25.379

    Law against the Manipulation of Public Opinion and the Popular Will (Anti-Troll Law)

    Directly related bill intended to prevent and sanction coordinated manipulation of public opinion through fake, automated or funded accounts, including algorithmic coordination and the use of artificial intelligence. It was filed in January 2026 and referred to the Legal Affairs Committee. In March 2026, the Supreme Electoral Tribunal objected to imprecisions in the text under consultation; no official evidence of a committee report or passage was found.

    In committee

This analysis presents evidence and gaps. It does not include public-policy recommendations: that conversation belongs to the country’s institutional actors.