Amplify Initiative: Building A Localized Data Platform for Globalized AI

2025-04-18

AI Summary

AI models exhibit poor cross-cultural adaptability due to overreliance on English and Western-centric data, resulting in deficiencies in relevance, utility, and safety—particularly in low-resource language regions. Method: We introduce an expert co-creation paradigm for multilingual data production, collaborating with 155 domain experts across five African countries to collect and collaboratively annotate 8,091 high-quality adversarial queries on the Android platform, covering seven indigenous Sub-Saharan African languages and focusing on sensitive topics such as misinformation. Our approach integrates cultural context, domain expertise, and data sovereignty principles, establishing localized data governance and authorship attribution mechanisms. Contribution/Results: This work yields the first high signal-to-noise adversarial query dataset tailored to African low-resource languages, enabling quantifiable evaluation of cultural adaptability and safety. It establishes a novel, equitable data infrastructure paradigm for global AI development.

Technology Category

• Natural Language Processing: Safety and Robustness
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Humans and AI: AI for Accessibility

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Current AI models often fail to account for local context and language, given the predominance of English and Western internet content in their training data. This hinders the global relevance, usefulness, and safety of these models as they gain more users around the globe. Amplify Initiative, a data platform and methodology, leverages expert communities to collect diverse, high-quality data to address the limitations of these models. The platform is designed to enable co-creation of datasets, provide access to high-quality multilingual datasets, and offer recognition to data authors. This paper presents the approach to co-creating datasets with domain experts (e.g., health workers, teachers) through a pilot conducted in Sub-Saharan Africa (Ghana, Kenya, Malawi, Nigeria, and Uganda). In partnership with local researchers situated in these countries, the pilot demonstrated an end-to-end approach to co-creating data with 155 experts in sensitive domains (e.g., physicians, bankers, anthropologists, human and civil rights advocates). This approach, implemented with an Android app, resulted in an annotated dataset of 8,091 adversarial queries in seven languages (e.g., Luganda, Swahili, Chichewa), capturing nuanced and contextual information related to key themes such as misinformation and public interest topics. This dataset in turn can be used to evaluate models for their safety and cultural relevance within the context of these languages.

Problem

• Addressing lack of local context in AI training data
• Improving global relevance and safety of AI models
• Co-creating multilingual datasets with domain experts

Innovation

• Leverages expert communities for diverse data
• Co-creates datasets with domain experts
• Uses Android app for multilingual annotations