Year: 2026
Type: report
Abstract: Unlawful by Design: Exposing the Human Rights Costs of Generative AI is a research briefing published by Amnesty International in May 2026 examining the development and deployment of generative artificial intelligence systems in relation to international human rights standards. The report outlines the technical and institutional context of generative AI, with a focus on data extraction practices such as large-scale web scraping used to train models. It situates these processes within legal frameworks protecting the rights to privacy, equality and non-discrimination, and freedoms of expression and thought. Drawing on examples of widely used AI systems, the briefing identifies systemic risks linked to the design of data pipelines, including non-consensual collection of personal data, reproduction of social biases, and environmental impacts.The report concludes with recommendations addressed to states, technology companies, and international bodies, including measures to regulate or restrict certain data practices, strengthen accountability frameworks, and align AI development with international human rights law.
Source: Issue Lab (Candid)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7164785708
