Rest of World

Rest of World is a New York-based nonprofit journalism organization launched in May 2020 by Sophie Schmidt, with a mission to cover technology's impact on people and communities outside the Western tech industry. Its AI coverage spans original investigations into content moderation labor exploitation, multilingual AI moderation failures, and AI-generated election disinformation across the Global South — establishing it as the primary English-language outlet for original investigative reporting on AI harms to communities in Africa, Asia, and Latin America.

Founding and structure

Schmidt, whose career included roles at Uber (public policy and communications), Xiaomi, and the Google-affiliated incubator Umbono, committed to spending as much as $60 million over a decade to launch and sustain Rest of World, with no advertising or subscription model. The organization incorporated as Rest of World Media Inc., a 501(c)(3) private operating foundation, in April 2019, and launched its first investigations in May 2020. A subsidiary — Rest of World Media International LLC — was registered in Kenya in January 2020 to employ editorial staff outside the United States. Anup Kaphle, previously a senior editor at The Washington Post and BuzzFeed News, joined as editor-in-chief at launch. By 2022, approximately 90 percent of its articles were reported by journalists on the ground in the regions they covered, drawing on a network of more than 250 freelance reporters and photographers. The Ford Foundation and Luminate are among its confirmed institutional funders; Luminate granted $200,000 specifically to expand AI coverage. In 2024, Rest of World received the Online Journalism Award for Topical Reporting (Small Newsroom category) and the National Magazine Award for design.

AI labor and content moderation investigations

Rest of World's most consequential original AI investigations document the exploitation of low-wage workers who build and moderate AI systems in the Global South. A 2023 investigation — "The workers at the frontlines of the AI revolution" — drew on interviews with freelance tech workers in Mexico, Nigeria, Pakistan, the Philippines, China, and India alongside a survey of 31,000 tech employees to document displacement and precarity facing content moderators, data annotators, and illustrators as AI capabilities expand. The investigation found that BPO companies projected AI could handle 20–30 percent of call volume by 2026, documented income drops of 33 percent among illustrators facing AI undercutting, and reported that the threat of AI replacement was used as a disciplinary mechanism to suppress wage negotiations among outsourced workers.

A 2025 investigation — "How Big Tech's AI labor supply chain relies on hidden African workers" — partnered with the African Content Moderators Union (ACMU) and the nonprofit Personaldata.io to map the presence of content moderators across 39 African nations performing outsourced digital labor for Meta, OpenAI, and Samsung through subcontractors primarily located in the UAE, North America, and Europe. Five ACMU moderators submitted data-access requests to their employers under data protection laws as an investigative methodology, revealing that companies systematically withheld information about the sharing of worker data with tech clients. The investigation documented wage theft, unsafe working conditions, psychological harm from exposure to graphic content, and nondisclosure agreements preventing workers from disclosing their conditions. A class action lawsuit filed in Kenya against Sama and Meta for worker exploitation followed the investigation's publication.

A 2023 investigation into AI content moderation and hate speech failures in Africa documented systematic failures of multilingual AI moderation systems whose training was dominated by English. The investigation drew on research from the Center for Democracy and Technology and examined how cross-lingual transfer limitations — essentially applying English-language AI moderation knowledge to African languages — produced structural gaps in detecting harmful speech in Amharic, Tigrinya, and other languages with limited digital representation. The investigation cited as a documented consequence the circulation of death threats against Ethiopian professor Meareg Amare on Facebook before his assassination in 2021 — a case in which AI moderation failure had a fatal outcome.

AI elections tracker (2024)

Rest of World created and maintained a 2024 AI Elections Tracker monitoring AI-generated content across more than 50 countries holding elections. The tracker drew on its global reporter network, which gathered and verified incidents throughout the year. A post-election analysis of 60 documented incidents from 15 countries found that AI was used more for memes and political campaign content than for mass coordinated deception — the "AI election tsunami" forecast by technologists and industry observers did not materialize. Notable documented incidents included AI-generated non-consensual intimate imagery targeting Mexican candidate Mariana Rodríguez Cantú, and Beijing-linked AI-generated disinformation videos about former Taiwanese President Tsai Ing-wen circulated one week before Taiwan's election. The tracker's Global South focus — covering India, Pakistan, South Africa, Indonesia, Sri Lanka alongside Western elections — filled a geographic gap in AI election harm documentation.

Role in the movement

Rest of World's Global South editorial mandate makes it the principal journalistic infrastructure for AI harm documentation in the communities most affected by AI systems they did not design and have limited recourse to contest. Its investigations on content moderation labor have anchored union organizing and civil rights advocacy among African AI data workers; the 2025 supply chain investigation was produced in direct partnership with the African Content Moderators Union. The AI elections tracker established a public documented record of AI disinformation across the Global South at a moment when comparative data was concentrated in Western elections. Rest of World's model extends the investigative journalism tradition of ProPublica and The Markup — data-driven original investigation of AI harms — into the geographic scope most absent from U.S.-based tech journalism.