Are humans at risk of becoming a “ghost in the machine” as artificial intelligence (AI) redefines good jobs? Generative AI is rapidly reshaping the way we live and work, but it’s also quietly affecting workers’ job security, financial stability and rights. Without significant intervention, we are entering an era of precarious work.
As generative AI adoption increases, more jobs – including white-collar ones – are at risk. The technology is remaking the human workforce while government, employers and educational institutions struggle to catch up.
“The changes are coming like a speeding train and we’re not ready for it. The level of disruption for the labour market is profound,” says Kevin Callinan, general secretary of public sector union Fórsa.
Although the recent focus on AI has concentrated on job losses, Callinan says it’s a bit more complicated.
RM Block
“It’s not simply a matter of if jobs disappear, it’s whether the work that remains gets better or worse. Whether it becomes more skilled, more autonomous and more secure or whether it becomes more intensified, more surveilled and more fragmented.”
Good jobs and precarious jobs used to be easier to define and measure. A good job had guaranteed hours, a pension, health insurance and other benefits. A precarious job was one with no contract or guarantee of work the next day.
The face of the precarious worker was typically female and found in the service sector (hospitality and food), among part-time workers or the solo self-employed.
Ireland’s Labour Force Survey shows that in the fourth quarter of last year, 582,500 people (20.6 per cent of everyone in employment) worked part-time. Of those, 20.8 per cent said they wanted more hours for more pay, but were unable to get them.
More recently, the precarious label has come to apply to those who use platforms and apps to get gigs – taxis, delivery drivers, carers – all of whom have no guarantee of work day to day.
Platform workers, even with recent changes to legislation, have fewer rights than employees, says academic Eddie Keane, of University of Limerick, who studies precarious work and the law. (An example of platform workers would include food delivery riders.)
He compares the case of the manager who works five days a week, has two days off and works 40 hours with that of a delivery driver who works five shifts in the week and has 260 “shifts” of work, all of which lasted for eight hours.
At the end of the year, the manager has one year’s continuous service. For the delivery driver, “because there’s no continuous service, there’s no unfair dismissal, no redundancy”.
When work is precarious, income levels vary and opportunities are harder to come by. You’re unlikely to get a mortgage, monthly rent can be difficult to meet and further education and training might be unattainable. You’re also likely to delay marriage and children.
Precarious work – no matter why it happens or who it happens to – can hurt self-esteem, reshape families, carve out communities and drag economies down.
Sustainable good jobs
The United Nations has just published its Preliminary Report of the Independent International Scientific Panel on AI, which says economic and labour gains from AI and their equitable distribution are not automatic unless there is investment in skills, workflows, infrastructure and labour-market institutions.
“Without these investments, AI risks widening inequality, displacing workers and shifting wealth from labour to capital rather than creating sustainable good jobs – those with fair compensation, worker autonomy and a reliable path to social dignity.”
Fair compensation is under threat if AI means workers lose out while business profits.
In this scenario, a growing proportion of the economy’s total income flows to the owners of assets – such as businesses, shares and property – rather than to workers’ wages. This is often illustrated through disparity between CEO pay versus average worker pay, now at its most extreme in the United States.
Worker autonomy may also be reduced.
“Increasingly, AI systems allocate tasks, evaluate performance, monitor behaviour and influence decisions about scheduling, pay or promotion. This means that workers can experience precariousness even when they have a permanent contract if they have little transparency over how decisions are made, limited ability to contest automated evaluations or diminishing control over their work,” says researcher Valeria Pulignano, professor of sociology at KU Leuven in Belgium, who studies work and AI.
The potential for human development is potentially at risk, too. “Measuring good work in the age of AI means asking whether technology enables workers to use and develop their expertise or whether it reduces their role to executing tasks defined and evaluated by algorithms,” Pulignano says.
When the machine decides what you do, how you do it and monitors your compliance, where’s the human dignity in that? You may as well just be a mechanical worker bee.
Slippery slope
Although workforce changes are happening quickly, our ability to monitor how AI is affecting the quality of jobs is unclear. Traditional indicators of job quality, such as the number of contract or part-time workers, don’t tell the full story.
In the age of AI, we need to pay closer attention to how work is organised, not just the terms of the employment contract.
AI is moving us from a job focus to a task orientation. Pulignano believes algorithmic management and supervision need to be included in our measurement of precarity.
“For example, the extent of digital surveillance, the degree of worker [human] autonomy over tasks and pace of work, transparency of AI-driven decisions, opportunities for human review and appeal, biases and whether workers can participate in decisions about the technologies that shape their jobs,” she says.
AI is redistributing decision-making power within organisations. Whether this strengthens or weakens workers’ rights depends on governance.
“If AI is deployed primarily to reduce labour costs and increase managerial control, it can contribute to precarious working conditions. If it is introduced with transparency, worker participation, human oversight and clear accountability, it can instead support better and safer work,” she says.
Given these transformations, two fundamental questions must guide our response, Pulignano says. These are: How can we ensure the equitable redistribution of the social gains of AI in the workplace? And how can we manage technological transitions in ways that prioritise job reconversion rather than displacement?
Good outcomes for workers and redistribution of the economic gains of AI will be a conscious policy choice, not a natural byproduct of the technology or the largesse of US multinationals.
Government needs to remember that AI is powered by data centres that use Ireland’s limited natural resources – our land, water, electricity – and they contribute to pollution, increased heat levels and climate change.
As such, AI providers and business users have financial, economic and social responsibilities here. Countries such as Norway have long insisted that revenues generated using natural resources – in its case, oil – are invested in the national budget, ensuring free university education, comprehensive universal healthcare and generous retirement pensions.
Irish industry is proactively advocating for upskilling over job losses. Last week, it launched a set of recommendations Skills for All, Skills for Life, a joint report from Ibec and Accenture.
If, as the report claims, “Ireland’s greatest natural resource and competitive strength has always been its talent” then we need to make sure no one – no matter how precarious their work – is left behind by AI.
Margaret E Ward is chief executive of Clear Eye, a leadership consultancy. margaret@cleareye.ie
