When we began putting together HRM, Artificial Intelligence and the Future of Work: Insights from the Global South, much of the public conversation about AI seemed to sit between two extremes. On one side was excitement about faster decisions, higher productivity and new kinds of work. On the other was anxiety about automation, surveillance and job loss. Both matter, but we felt an important question was being missed: whose future of work are we talking about?
Too often, debates about AI at work assume the infrastructure, labour markets, institutions and regulatory systems of wealthier economies. Yet AI does not arrive in a vacuum; for example, a recruitment tool, platform algorithm or HR analytics system can have very different consequences where internet access is uneven, social protection is limited, informal work is widespread, or regulation is still developing. The book grew partly from our concern that discussions of AI and HRM needed to take these differences much more seriously.
That was why we brought together research spanning Asia, the Middle East, Latin America and Africa, examining issues ranging from recruitment and HR analytics to gig work, digital leadership, higher education, tourism and the ethical use of AI. Across these very different settings, one message became difficult to ignore: the future of work will not be determined by technology alone. It will be shaped by the choices organisations, governments and workers make around it.
More work is not the same as decent work
Regardless of the benefits of AI and its growing ability to create or widen access to economic opportunity, access to work is only part of the story. The same technologies can also intensify existing inequalities. In platform work, for example, one of our earlier research suggests that algorithms can influence who receives work, how performance is judged and what workers earn. Yet these decisions may be difficult for workers to see, understand or challenge. People may gain flexibility while also facing insecurity, limited social protection and weaker bargaining power. Our book highlights precisely this tension in its discussions of platform work and algorithmic management.
This matters directly for Sustainable Development Goal 8. SDG 8 speaks beyond simply creating more jobs to reinforcing the need for productive employment and decent work. If AI expands employment while leaving people with unpredictable incomes, opaque decision-making, poor working conditions or little protection, then technological progress and decent work are not necessarily moving in the same direction.
So, the question cannot simply be, “How many jobs will AI create or replace?” We also need to ask: What kind of work is being created? Who benefits from it? And who carries the risks?
The Global South is not simply a place that needs to “catch up”
A second lesson from our book is that responsible AI cannot mean importing technologies and practices developed elsewhere and assuming that they will work in exactly the same way.
Many AI systems are developed within contexts that have different languages, data, workplace norms, infrastructure and legal protections from those in which the technologies may eventually be used. Therefore, we raise concerns about technologies developed predominantly in the Global North being transferred without sufficient attention to the economic, political and sociocultural realities of the places in which they are deployed.
Localising AI matters, meaning that we need to start asking whether technologies fit local needs and values, whether the data they rely on adequately represent the people affected by them, and whether organisations have the infrastructure and skills to use them responsibly.
This is not an argument for slowing innovation. It is an argument for making innovation useful. Investment in digital infrastructure, research capacity and skills is therefore just as important as investment in AI itself. Without reliable connectivity, affordable access and opportunities to develop new skills, the benefits of AI are likely to remain unevenly distributed.
HR cannot be brought in at the end
One of the strongest implications for us as HRM scholars concerns the role of the people function.
AI in the workplace is often treated primarily as a technology decision. But once a system affects recruitment, performance, scheduling, pay, learning or employee data, it is also a people decision.
HR therefore needs to be involved early enough to shape how technology is introduced, rather than simply being asked to explain or manage the consequences afterwards. This is particularly important because HR professionals have responsibilities around employee well-being and fairness, even though they are not always centrally involved in decisions about technology investment and implementation.
That means asking practical questions about job redesign, skills, fairness, privacy, worker voice and well-being. It also means creating routes through which employees can question or challenge decisions influenced by algorithms.
Aligning SDG 8 with AI transition
Ensuring that AI makes decent work better is a responsibility for all. For governments, it means developing AI strategies that treat worker protection as part of innovation rather than as an obstacle to it. Regulation needs to address transparency, privacy, fair treatment and oversight of algorithmic decisions. The absence of mature regulatory frameworks in parts of the Global South makes this especially important.
For employers, it means judging AI investments not only by cost savings or speed, but also by what happens to job quality, autonomy, workload, learning and employee voice.
For education providers and policymakers, it means making reskilling and digital learning accessible rather than assuming that workers will somehow adapt on their own. And it means ensuring that countries in the Global South have a meaningful voice in international debates about AI governance, rather than simply receiving technologies and rules designed elsewhere.