Register hereWorld Bank Group
a) Smart Greenhouse
A prototype of a low-cost, offline-first agricultural intelligence system that combines in-greenhouse IoT environmental sensors with a small AI model running entirely on a smartphone. The system operates without any internet connection, translating real-time local conditions—such as temperature, humidity, and barometric pressure—into simple, actionable guidance, and leverages on-device RAG with curated agricultural knowledge to deliver targeted, context-aware recommendations tailored to the specific greenhouse environment.
b) Compliance Agent
An AI-powered tool that helps financial regulators and institutions ensure that financial service contract templates, such as loan and credit-card agreements, meet regulatory and legal standards. Acting like an intelligent clerk, it reads contract templates, flags potentially non-compliant clauses, links each finding to the relevant legal provision, and suggests corrective wording in seconds; the platform serves supervisory authorities reviewing submissions and financial institutions self-assessing contracts before submission, while human reviewers remain in control with every decision tracked and auditable. Multilingual and adaptable across jurisdictions, the Compliance Agent is scalable for strengthening regulatory efficiency and consumer protection across the financial sector, particularly in emerging markets.
c) GeoLab
GeoLab stands at the forefront of geospatial innovation, providing World Bank Group operations with state-of-the-art AI-powered tools, fostering collaborative knowledge exchange, and focusing on targeted capacity development. At the AI for Good conference, GeoLab will showcase an interactive suite of water-related datasets—including precipitation patterns, drought indicators, water stress metrics, and flood hazard assessments—that demonstrate the transformative role of geospatial AI in tackling complex, real-world issues. Attendees will explore how GeoLab’s solutions can map surface water distribution, analyze seasonal trends to enhance irrigation strategies, and evaluate flood risks to support smarter environmental planning, boost agricultural productivity, and strengthen links to employment opportunities, highlighting how advanced geospatial data supports more informed decision-making.
d) Development Data Partnership AI Language Content Library Program
The team will demonstrate the Development Data Partnership AI Language Content Library Program through a video and voice-based phone application. Most AI language models work well in English and a handful of other widely spoken languages—but billions of people speak languages with almost no digital training data. This Gates Foundation-supported initiative licenses proprietary audio, text, and structured data from broadcasters, publishers, and community organizations in underrepresented languages to close the training data gap that prevents high-performing AI models from being built for these languages by making datasets available to vetted researchers for non-commercial NLP, ASR, and LLM fine-tuning work.
e) Trustworthy Dissemination of Development Data
The team will demonstrate the “Trustworthy Dissemination of Development Data” project — showcasing how the World Bank is using AI to make development data easier to access, understand, and trust. The solution uses the Model Context Protocol (MCP)—a standardized way for AI systems to securely access authoritative data—and Proof-Carrying Numbers (PCN), which allows users to verify that numerical results come directly from trusted World Bank data sources. At the core is the Data360 MCP server, which acts as a bridge between AI applications (such as chat tools) and official development data and metadata. This enables policymakers, researchers, and the public to ask questions in natural language and receive answers grounded in verified data, with clear attribution and transparency. Target users include World Bank teams, government counterparts