Fairwork
Ratings
Country Ratings by Region
Total points awarded across 337 platform ratings
Cloudwork Ratings
Cloudwork is work that can be performed remotely via digital work platforms.
The Fairwork Cloudwork Report 2022 assessed and scored basic standards of fairness in working arrangements on 15 online remote digital labour platforms, according to our five Fairwork principles. The platforms studied were selected based on their global reach (e.g. Upwork, Amazon Mechanical Turk, and Scale/Remotasks), their position as market leaders (e.g. Workana), and companies focused on specific segments, for instance, design (99designs) or academic research (Prolific). In addition to the desk research on the platforms and conversations with managers, we conducted a worker survey between January and July 2022. For the global platforms we sampled up to 60 workers per platform, with a relatively even distribution of workers by continent, and for the regional platforms, we sampled at least 15 workers per platform from that region. After cleaning the data, we were left with responses from 613 workers in 84 countries.
In general, cloudwork platforms are still not close to safeguarding the basic standards of fair work expressed in our five principles. One platform scored seven points (Prolific) out of 10, one scored five (Jovoto), and one scored four (Workana). For the other 12 platforms, we were unable to evidence that they met more than three of our ten thresholds. For four platforms, we could not find evidence that they met any threshold.
Our survey revealed that, on average, workers spent over 8.5 hours per week on platforms on unpaid tasks. These include searching for clients or tasks, applying for jobs, building or curating online profiles, submitting work to competitions, taking unpaid qualification tests, and dealing with overly demanding clients. In addition, around a third of our respondents reported the experience of completing a task that they were then not compensated for.
Fairwork AI
From managing productivity to determining which candidates get hired, AI systems are having profound effects on our daily routines. But these effects are not universally positive. The risks associated with AI in the workplace range from reductions in job quality and spikes in work intensity to algorithmic discrimination and ubiquitous surveillance. For many workers, the introduction of AI systems to the workplace leads to extremely unfair outcomes. So far, the debate around the ethics of AI has generally skipped the question of work, with the debate instead either focusing on the risks AI poses to society as a whole or to individuals in their role as citizens1. But a new Fairwork research stream funded by the Global Partnership on Artificial Intelligence (GPAI) aims to change that.
Drawing on the Fairwork project’s research on digital labour platforms, Fairwork AI, based at the Oxford Internet Institute, have developed a set of ethical principles and associated benchmarks to guide the deployment of AI systems in the workplace. These principles build on the OECD Recommendations on Artificial Intelligence2 and were generated through a two-stage multi-stakeholder consultation with representatives from trade unions and worker representatives, governments and quasi-governmental organisations, and academics and experts including the International Labor Organization, Uber, Microsoft, the International Transport Federation, and the UK Information Commissioner’s Office, among others.
We completed two rounds of engagement with these stakeholders to understand their perceptions of the draft principles and engage with them on questions we felt we had yet to adequately address. The first round involved 21 in-depth interviews with a range of stakeholder groups, and a focus group to understand the convergences and divergences in their perspectives. The second round of engagement consisted of a survey that presented participants with sections of the penultimate draft of the principles. This resulted in 117 responses. We conducted a thematic analysis on the resulting data and then designed a series of edits to the principle text, before publishing a report with GPAI, published in November 2022.
The Fairwork AI team is now leading an extended impact phase to put these principles into practice. While more and more concrete legislative action is being taken to regulate AI systems, from the EU AI act to the US Algorithmic Accountability Act, our experience in the platform sector shows that regulatory action can benefit from non-statutory, civil society-led monitoring and standard-setting approaches such as those undertaken by the Fairwork project. As AI regulation begins to be developed across the world, we need multinational monitoring of work conditions with a consistent and cross-comparable methodology, and second, the creation of a set of practical standards of fairness and a system that applies scrutiny as a way to leverage private sector actors to proactively make change – thereby demonstrating the feasibility of fair work.
Our Fairwork methodology has opened up space for an ecosystem of policy actors to understand the current state of play, and take meaningful action to mitigate the risks of the platform economy in a way that complements the development of concrete legislation. Going forward, we will be highlighting the fundamental questions of fairness posed by the widespread deployment of AI in the workplace, providing information on the existing practices, risks and outcomes of this deployment, and helping to shape the standards through which this deployment is evaluated. Throughout, Fairwork remains fundamentally committed to understanding and amplifying workers’ experiences of work, as a primary step towards enabling fairer outcomes.
Built with Shorthand