Artificial Intelligence
Given the profound and rapid technological changes shaping our world, a common theme discussed throughout the conference was the role of AI in shaping and reshaping labour markets. These discussions included how technology and AI can promote fair work and strengthen productivity, while addressing risks. Participants also explored how to harness AI for growth, while addressing the uneven distribution of its benefits.
Participants stressed that the extent of change amounts to a fourth industrial revolution. While AI and automation offer significant opportunities for productivity gains and innovation, they also pose challenges for employment stability, including the risk of large-scale displacement in roles with high routine content and a potential decline in job quality. It is crucial to continue to constantly monitor and debate the role of AI in the world of work, even in the absence of full information.
There are some misconceptions around the role of AI in labour markets. These include:
- The assumption that AI automatically causes job losses. This presumption risks normalising and accepting job losses as inevitable. This is not to negate the fact that some jobs will no longer exist in the same way as they previously did, and that there will be job displacement. These impacts should be managed.
- The misconception that the use of AI and its impacts will be uniform. The use of AI varies significantly across and within different labour markets. The same is true for the impact that AI will have on different labour markets and different types of workers, businesses, unions, and governments. Generalisations can be unhelpful and result in poor policy making.
- AI’s impact will be beyond human agency. There should be a feeling of empowerment with AI, which can be controlled to ensure positive impacts on the world of work. AI will alter the demand for jobs and will change the world of work, but this is not necessarily bad – it is up to us to decide how to use AI.
The use of AI in workplaces is rapidly spreading across all sectors and can boost productivity at both a micro- and macro-level, but these productivity gains require distributional consideration. Deploying AI in workplaces can enhance workers’ performance. In the UK, it is estimated that around 70% of workers are in occupations where AI could perform or enhance tasks[1]. This may improve the quality of jobs, if workers can use the time saved by AI technologies to work on meaningful tasks. However, there is a risk that this leads to intensified workloads and job degradation.
On a macro-level, AI is a driver of innovation and can lead to the creation of new jobs. AIwill remove parts of jobs that are easily automated and can restructure labour markets in a more productive way. However, this positive result only happens when innovation is strong, and AI is adopted by employers in a way that increases demand for skills and jobs.
AI will likely impact specific groups in the labour market more than others and employment rights should be updated accordingly. In many countries, women are over-represented in jobs that involve tasks that AI can replace[2]. This over-exposure to the negative impacts of AI is compounded by the fact that women are less likely to receive AI training and, in some countries are generally more sceptical of using AI. This is accompanied by a trend towards more restrictive gender norms. There are also barriers to adoption among some groups e.g. older women and some organisations, partly due to different organisational cultures around AI and due to uneven learning curves amongst workers. This increases the importance of training workers to build confidence and capability to effectively use AI to enhance the quality of their jobs.
The adoption of AI in workplaces will impact the development of skills for those who are early on in their careers. There are some tasks that, despite being routine or basic, are necessary foundational skills to build – particularly in a knowledge economy. Care needs to be taken to ensure that workers can develop these skills amidst the increased automation of routine tasks.
Algorithmic management is increasingly being used across all sectors, and is not limited to the platform economy, raising significant questions about surveillance and autonomy at work. Using algorithms to set pay, monitor workers, and recruit, can lead to negative unintended consequences such as opaque decision making, intrusive monitoring and biased recruitment systems. There is also asymmetry of information between workers and employers as to how AI technologies are being used.
There are wider risks of deploying AI without proper oversight. This includes concerns about data rights e.g. how worker-generated data is used and protected and the wider environmental impact of using AI.
Policy makers should therefore consider the benefits of AI in workplaces, and what role governments, businesses, unions and workers should have in securing fair distribution of these. Government plays a role, but exactly what this is depends on national and local contexts. Regulation can ensure that employment rights frameworks are updated for AI-driven workplaces.
Recommendations
Given the nature and pace of change, interventions should:
- Consider how to distribute the productivity gains of AI fairly to support both economic growth and improved outcomes for workers.
- Respond rapidly to the sheer pace of change. Interventions should be quick and flexible.
- Centre transparency and agencyfor workers. This is essential to ensuring that workers are brought along throughout the AI transformation, and that changes do not simply happen to them. Trade unions can play a role in facilitating this.
- Focus on building trust and dialogue between workers and employers. As AI technologies are introduced into workplaces, and as they continue to evolve, workers and employers should be in continuous dialogue to ensure AI is used fairly and transparently. As use of algorithmic management increases, stronger trust building measures are required in recruitment processes and workplace monitoring. Examples of interventions suggested by participants included co-created AI strategies and policies, and worker led choices to technological changes.
- Create clear governance frameworks that explicitly incorporate AI. This should include transparency around acceptable and unacceptable uses.
- Account for regional variations in the use and impacts of AI. Hyper-local interventions can be developed. Other interventions could include job guarantees, especially in managing communities through industrial challenges/transitions.
- Involve gathering and analysing data. There are currently different indices and proxies that are used to measure the impact of AI. These are often inaccurate and result in businesses making significant decisions on AI strategies based on general assumptions. There should be greater examination of what metrics are useful, and who should be collecting this data. For example, there should be national statistics that measure job qualityto better understand the impact of AI on the quality of work.
There are broader questions of the impact of AI on the concept of work and how it is shaping the relationship with work. Policy makers need to consider the role of AI in reshaping the meaning of work, jobs and our societies. The way in which AI is used can determine whether work becomes intensified or more flexible and can allow jobs to be designed in a different way. A humanist approach to AI holds human resilience as central to this progress, considering psychological, cultural, environmental resilience and not just economic resilience.
Platform Economy
Digital labour platforms are creating new markets for businesses and expanding job opportunities for people, including those who would otherwise be unemployed. However, these digital platform models challenge established workers’ protections. In this session, participants explored how to ensure fair working conditions in the platform economy while continuing to support innovation and growth. Participants discussed definitions of platform work, its potential benefits and risks, and recommendations to address these.
Defining platform work is complex. Platform work encompasses paid work, organised or mediated through technology (website, application), which matches demand for a service with an individual who can supply this. It uses algorithmic systems to allocate tasks and sets conditions to monitor performance. The platform itself does not define employment status. However, whether platform work can be categorised as a sector is up for debate. On one hand, platform work has distinctively different features and different labour markets – suggesting it is a unique sector. On the other hand, platform work is not limited to a single service / sector in that it provides a service across sectors through technology by matching the demand of a service and supply of workers. The number of tasks available to workers and frequency of these tasks is regulated by algorithms. Platforms supply services beyond typical sectors such as food or transport and include sectors such as health care and education. These platforms should have a responsibility towards those working through their technologies.
Platform workers are not limited to one type of employment status category or one model of employment. Platform workers can be employees, workers or self-employed. Their employment status will depend on different factors and can change.
Platform work has numerous benefits, including:
- Diverse forms of work which result in flexibility for workers. This gives people greater individual freedom to choose how and when they work. People can engage in different types of jobs, including taking up several jobs at the same time. Platform work provides geographical flexibility as workers can often choose where they take up work. In Ukraine, platform work offers the labour market some stability in an otherwise unstable environment, where they have lost a quarter of their workforce. In Ukraine, platforms provide people, especially young people, with the ability to earn money quickly, to gain work experience, and to earn money from different and/or multiple jobs at the same time.
- Employment opportunities for those who might otherwise be excluded from the labour market. The flexibility provided by platform work attracts people to work who otherwise would not be able to undertake traditional employment (reducing inactivity). It also allows for quick access to work – allowing people to quickly earn additional income or to easily gain access to work for a limited period i.e. more flexible form of part-time work.
- Broader economic opportunities. It matches demand for services with a large (and often international) supply of skills, quickly.
Participants also identified risks from platform work including, but not limited, to:
- Undermining of employment rights due to power imbalances between employers and platform workers. Platform providers have greater power to set conditions for workers compared to traditional work. Employment rights could be impacted, e.g. with workers experiencing increased vulnerability and insecurity, poor health and safety protections (partly because this type of work is difficult to inspect), and in the worst cases, risks of modern slavery.
- Data issues, such as the portability of ratings across platforms. Platform workers’ reputation exists solely on, or is owned by, the platform. As consumer ratings are vital for businesses particularly SMEs, the question of who owns and manages this data determines the power dynamic between the platform provider and the individual.
Recommendations
Platform work discourse should centre on addressing power imbalances and increasing job quality. As platform providers have greater power to set conditions compared to traditional work, there should be greater focus placed on the responsibilities of platform providers.
Given that platform work can span across different sectors, demographics of workers and geographical borders, interventions require tailoring / cannot be uniform. International examples include the European Union’s Directive on Platform Work, Spain’s “Ley Rider” (Rider Law), (Spain was the first country to issue regulation in the sector), in 2021) and India’s Code on Social Security (recognition of the platform economy and broader protection for platform workers, in effect from November 2025).
Other suggestions included:
- Greater transparency in discussions around platform work. There are parallels between the discourse around social media and its impacts, and the technology companies that facilitate platform work. Platform workers should be made aware of how algorithms are being used to manage their work.
- More agency for platform workers. This could include giving platform workers more control over algorithmic work, for example, clearer choices and mechanisms to challenge decisions made by algorithmic managements or to request meaningful human oversight. Another suggestion was to create a genuine mechanism for platform workers to opt in or out of algorithmic systems.
- Human-centred and transparent approach to algorithmic management and automated decision making. Platform workers should understand how algorithms operate and impact their work. This transparency should extend to consumers, e.g. if companies use surge pricing, consumers should be made aware of the surge paying of employees.
- Clarity on employment status. Platform workers should have the correct employment status. One option could be using the ‘worker’ status as a default, but businesses may disagree with this approach. Another option could be giving those who are genuinely self-employed similar protections to those with other employment statuses. Ultimately, there should be a practical and workable solution that does not rely on presumption as this will lead to unnecessary strain on the enforcement system.
- New forms of collective bargaining / trade unionism to adapt to the new type of work that platform work represents.
- Access to welfare: Providing welfare provisions to platform workers, such as insurance and social security for platform workers.
- Consumers should also be part of the conversation about how technology, algorithms and platform work shape the quality of work and services provided. Firstly, platform workers and consumers are not always separate entities. Secondly, there should be consideration of protections for consumers e.g. in relation to dynamic pricing.
Given that platform work transcends national borders, it is crucial to discuss the role of regulation at an international/global level. This could include regulation at a regional level such as the EU’s Platform Work Directive. To ensure more coverage globally, this should also include regulation at a multilateral level, for example, as is currently being negotiated at the International Labour Organization.
The impacts on the wider labour market should be considered when designing and implementing regulations. For instance, the regulation of platforms may change the supply of labour in different markets.
It will be challenging to measure the impacts of regulations or interventions on the platform economy. Data is not comprehensive as to the size, gender and geographic breakdown of the platform economy. This makes it harder to anticipate how firms might respond to regulation. It will take time to get the regulation right, and new regulation may not always be necessary, e.g. where technology facilitates existing practices, existing labour laws already apply.
[1] International Monetary Fund, ‘Gen AI: Artificial Intelligence and the Future of Work’, 14 January 2024, Link: https://www.imf.org/en/publications/staff-discussion-notes/issues/2024/01/14/gen-ai-artificial-intelligence-and-the-future-of-work-542379
[2] International Labour Organization, ‘Gen AI, occupational segregation and gender equality in the world of work’,5 March 2026. Link: https://www.ilo.org/publications/gen-ai-occupational-segregation-and-gender-equality-world-work