In a report presented at the Global Partnership on Artificial Intelligence (GPAI) Summit in New Delhi on 12 December 2023, the Fairwork project has shown the impact the application of its fair working principles can have on workers operating behind the scenes in the artificial intelligence (AI) industry.

The deployment and development of AI relies on people engaged in poorly paid, repetitive work. This work is often focused on making sure that the information AI systems use is correctly tagged and labelled. This type of work, primarily performed in the Global South, is called ‘ghostwork’ as it is conducted by people but widely believed to be an automated process.

The Fairwork project at the Oxford Internet Institute, University of Oxford has completed its first report into the working conditions of AI workers. The report focuses on Sama, a key node in the global production networks of AI that provides data annotation services for some of the major global internet companies. As a certified BCorp with a reputation for being ethical, Sama bases production in East Africa, specifically seeking to hire people who live below the poverty line and lack formal employment opportunities, a process called “impact outsourcing”.

Through desk research, company management interviews and worker interviews, Fairwork assessed Sama against the Fairwork AI Principles, which cover pay, conditions, contracts, management and representation, and companies can get a maximum score of 10/10 based on the peer-reviewed Fairwork research methodology. Fairwork fieldwork initially identified several concerns; such as precarious contracts, excessive overtime, dangerous levels of job strain and discriminatory management practices. One worker explained how she was working seven days a week from 7.40am to 6pm. She hadn’t received overtime pay for three months. Sama acknowledged Fairwork’s findings and committed to working with Fairwork team to improve working conditions. At the end of the first assessment, Sama was awarded 5/10 for meeting the requirements for Fairwork principles of fair pay, conditions, as well as the first threshold of fair contracts. The report explains:

“Our ongoing engagement with the company throughout the research process demonstrated yet again that significant positive changes are possible, and that these changes can have a huge impact on the lives of workers, be they workers for platform companies, or workers involved in the global production networks that produce AI technologies.”

Dr Funda Ustek Spilda, Senior Researcher and Project Manager at Fairwork explained the impact of the report: “It’s not acceptable that workers in the AI industry are subject to working conditions that put their health, wellbeing and financial stability at risk. Throughout our research Sama was very open to granting access to Fairwork researchers, hearing our findings and thinking about ways to improve the working conditions for their workers. Our engagement shows that companies can make meaningful changes in the industry, and together we can raise the tide for all AI workers.”

Notes for Editors:

The full report can be accessed here.  The report was funded by the Global Partnership on Artificial Intelligence (GPAI) and peer-reviewed by Jonas Valente, Researcher, Oxford Internet Institute and Rafael Grohmann, Researcher, University of Toronto.

About Fairwork

The Fairwork project is coordinated from the OII, University of Oxford, and the WZB Berlin Social Science Centre. The aim of the Fairwork project is to highlight best and worst examples of how new technologies are being used in the workplace, in order to show that better and fairer jobs are possible in the digital economy. The Fairwork AI report was developed in the context of the ‘AI for Fair Work’ project, which was launched in 2021 by GPAI: Global Partnership on Artificial Intelligence. In 2023, the GPAI formally partnered with the OII-based Fairwork project to develop a scoring framework that can be used to examine the impact of AI on workplace fairness. The research for the report was generously supported by GPAI and conducted in collaboration with experts of the GPAI Future of Work working group. The project is led by Professor Mark Graham at the University of Oxford. The report was authored by Callum Cant (University of Essex, previously Oxford Internet Institute), Funda Ustek Spilda (Oxford Internet Institute) and Lola Britain (Oxford Internet Institute and Mark Graham (Oxford Internet Institute).

About Oxford Internet Institute

The Oxford Internet Institute (OII) is a multidisciplinary research and teaching department of the University of Oxford, dedicated to the social science of the Internet. Drawing from many different disciplines, the OII works to understand how individual and collective behaviour online shapes our social, economic and political world. Since its founding in 2001, research from the OII has had a significant impact on policy debate, formulation and implementation around the globe, as well as a secondary impact on people’s wellbeing, safety and understanding. Drawing on many different disciplines, the OII takes a combined approach to tackling society’s big questions, with the aim of positively shaping the development of the digital world for the public good. https://www.oii.ox.ac.uk/

For more information or to speak to a member of Fairwork please contact:

Roz Pacey/Sara Spinks Media and Communications Manager

T: 01865 287237 E: press:oii.ox.ac.uk