Fairwork AI

Language:

Understanding the Labour Behind AI

There is nothing “artificial” about the immense amount of human labour that builds, supports, and maintains AI systems. From data annotation and content moderation to model evaluation and warehouse logistics, millions of workers contribute to the development and deployment of AI technologies.

Fairwork investigates the labour that underpins AI systems through independent research, worker engagement, and evidence-based analysis. In doing so, our work has documented the conditions of AI data work, exposed hidden AI supply chains, and identified the governance challenges associated with increasingly fragmented forms of digital labour.

Through independent research, worker engagement, and evidence-based analysis, Fairwork makes visible the labour practices, risks, and inequalities that underpin today’s AI economy.

What We Study

Fairwork’s research examines how AI systems are built, maintained, and governed, with a particular focus on the workers who make these systems possible.

Our work explores:

Data annotation and model evaluation
Content moderation
AI supply chains and outsourcing
Platform work and digital labour
Labour governance and regulation
Worker voice and representation
Responsible AI policy and practice

By combining worker engagement, company evaluations, policy analysis, and field research, Fairwork contributes evidence to debates on labour standards, AI governance, and the organisation of digital work.

What Our Research Shows

Through its research, Fairwork has:

Documented the conditions of AI data work, including data annotation, content moderation, model evaluation, and related forms of digital labour
Exposed hidden AI supply chains, making visible work that is often outsourced, geographically dispersed, and difficult for clients and regulators to trace
Identified labour risks across digital labour systems, including low pay, insecure contracts, opaque management practices, and limited worker voice
Shown how fragmented outsourcing can create governance gaps across clients, platforms, suppliers, and subcontractors
Developed the Fairwork Principles as a framework for assessing fairness in digital labour and improving labour standards across sectors

Together, these findings contribute to a clearer understanding of how AI systems are produced, where accountability breaks down, and what is required to improve the conditions of the workers who power them.

The Fairwork AI Principles

One of Fairwork’s core contributions has been the development of the Fairwork Principles, a globally recognised framework for evaluating fairness in digital labour.

Developed through extensive consultation with workers, researchers, policymakers, trade unions, and industry stakeholders, the Principles have become an internationally recognised benchmark for understanding and evaluating labour standards in the digital economy.

Today, they inform Fairwork’s research, ratings, policy engagement, and evaluation activities across multiple sectors and regions.

Publications & Reports

Fairwork publishes ratings reports, case studies, policy briefs, and academic research examining labour conditions across digital platforms and AI supply chains.

Our research has been conducted in partnership with organisations including the Global Partnership on Artificial Intelligence (GPAI), Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ), governments, trade unions, civil society organisations, and academic institutions.

Together, these publications contribute to a growing evidence base on digital labour, AI supply chains, labour governance, and the conditions of the workers who power AI systems.

From Research to Practice

Fairwork’s research has informed the development of an independent evaluation and certification framework for organisations seeking to assess and improve labour conditions across AI supply chains.

Grounded in the Fairwork Principles and informed by worker experiences, this framework translates research into practical tools for understanding labour risks, strengthening accountability, and supporting more responsible AI governance.