Who will benefit from AI in the workplace and who will lose out?

The advent of generative AI, such as ChatGPT, is potentially a seismic shift for work and organisations.

For the last two years, we have been collaborating with the Institute for the Future of Work on a research project that asks precisely this question.

AI is not just replacing physical and cognitive tasks, it is also used in managing people – sometimes called ‘algorithmic management’ or ‘bossware’.

This has the potential to improve productivity by making decisions more quickly and accurately, combining diverse information without human co-ordination, and identifying patterns in consumer behaviour to make accurate market predictions that previously relied on subjective judgements.

However, it can reduce workers’ autonomy, agency, satisfaction, and wellbeing.

Numerous studies have offered dire warnings about widespread job destruction, estimating that between 10 per cent and nearly half of occupations face potential disruption.

When it comes to the nature of the disruption, the research has been rather binary. Will jobs be created, or disappear?

Yet there are important nuances. Both outcomes can occur at the same time. More pertinent is what types of jobs are created and eliminated. Which skills are becoming obsolete and which must be developed?

Finally, what is the nature of the work that remains? Is it enhanced by technological augmentation or is job quality diminished?

On one hand, autonomous machines can automate dull and repetitive mental tasks or physical tasks that are dangerous or unpleasant, freeing up human workers to engage in more interesting and meaningful work. This technology can also augment human capabilities, enabling us to perform similar functions more effectively.

These scenarios have significant benefits for work and wellbeing.

Which workers are at greatest risk from AI?

On the other hand, AI could be used to de-skill work, lower pay, make employment more precarious, diminish the social aspects of work, and reduce opportunities for personal development.

In our research, we often see evidence of the skill-enhancing, job-creating benefits of AI and the negative impacts in the same context. So who is benefitting and who is losing out?

There are three major dynamics at play. The first is called skills-biased technological change. This is where new technologies create a demand for higher skills. Those with the necessary skillset, or the capacity to acquire those skills, will be in high demand and benefit from the adoption of AI.

If these skills are widespread, the benefits will be widely felt in the economy. For much of the 20th century, skill-biased technological change was complemented by an ever-expanding educated workforce, especially in the US economy. The rising tide lifted all boats.

However, in the final quarter of the 20th century, the skills advances tailed off and less of the workforce was able to benefit from technological change, driving greater economic inequality. A small slice benefitted from ever-rising pay, while a significant chunk lost ground, especially in terms of real wages.

AI will have similar skills biased effects. Where there are insufficient skills in the labour market, it will be hollowed out, eliminating semi-skilled work and driving a significant portion of the workforce into lower-paid activities that have yet to be automated.

A lucky few, able to invest in their skills, will have their jobs and personal skills enhanced, and will benefit economically. Those with the highest skill have, until now at least, been beneficiaries. However, AI is so pervasive that it is likely to creep further up and down the skills continuum, further disrupting highly trained professions, as well as helping to automate administrative work.

How leaders can help workers benefit from AI

The second major dynamic builds on the first. Recent research shows the impact that adopting robots has on work depends very much on the institutional environment, and in particular the investments a country has made in its ‘innovative capabilities’. This includes investments in human capital, but also in research-focused institutions, and other scientific establishments.

For countries that invest in their national innovation system, the effect of new technology is much more beneficial: robot and human labour are treated as complements, thus work and outcomes are more positive. But in countries that fail to invest in this innovation system, the technology is much more likely to substitute for human labour.

Our research provides evidence that this environmental impact occurs at a regional level as well. Areas of the UK with greater investments in human capital and technical infrastructure, such as high-speed internet connectivity, are better prepared to adopt technology, so jobs, skills, and quality of work are more likely to be increased rather than eliminated. National and local government action is essential to ensure that we have an educated and innovation-ready workforce. Under such circumstances, we predict the balance of impacts of AI will be more positive than negative. Failing to ‘level up’ those regions that are less ‘innovation ready’ will see even greater inequality as a result.

The third major dynamic is how we choose to manage our human resources. The significant benefits of taking an investment orientation to managing people have been recognised for decades. Increasing skills, rewarding performance, providing greater autonomy, sharing information, and other forms of workforce engagement have been shown to deliver positive results in a wide range of contexts.

Yet adoption of such advanced management practices is far from universal. Global comparisons have shown the UK lags behind significantly in adopting productivity-enhancing management practices, especially among SMEs, with consequences for national productivity.

These same practices are associated with the capacity of organisations to adopt new technologies. They also impact the outcomes – in terms of enhancing job creation and the quality of the jobs that are created. Organisations that share information with their workforce, that train employees, that consult with employees about major issues, and that empower their employees with higher levels of autonomy and responsibility, are more likely to take advantage of AI in ways that complements their workforce rather than making it redundant. Potential applications of AI are more likely to be identified by employees in the front line or the middle tier of the organisation.

It is easier to adopt new technologies with a collaborative workforce that has the skills, motivation and opportunity to leverage it. We find that, all else equal, an investment approach to managing people has a positive benefit when it comes to the adoption of AI and the consequences of that adoption.

Regardless of what regulation or legislation eventually emerges, we can expect that those employers that pay attention to these risks and impacts, and that engage with their workforce proactively and collaboratively, not only in adoption, but also in evaluating the impacts of AI on work and worker wellbeing, will be best placed to realise the benefits that are possible, while protecting themselves from the risks that are posed as we adjust our organisations, business models, and workforces to what is arguably the greatest technological shift in the last 50 years.

Professor James Hayton's research focuses on how human resource management practices foster organizations' capacity for entrepreneurship and strategic renewal. He teaches on the Executive MBA and the Global Online MBA, as well as on the PhD Business and Management.  

For more articles on the Future of Work, and Digital Innovation and Entrepreneurship, sign up to the Core Insights newsletter here.