AI creates both inclusion and exclusion in labor markets

The impact of artificial intelligence (AI) is proving far more complex than simple narratives of job creation or job loss. A new systematic review finds that AI is simultaneously expanding opportunities for inclusion while deepening risks of exclusion and discrimination across workplaces.

The study, titled “Artificial Intelligence in the Labor Market: Evidence on Worker Inclusion, Exclusion, and Discrimination—A Systematic Review,” published in Sustainability, examines how AI is reshaping employment access, job quality, and workforce dynamics. The findings suggest that AI is not inherently inclusive or exclusionary. Instead, its impact depends heavily on how it is designed, deployed, and governed.

AI is restructuring jobs through automation and task reallocation

The study shows that AI is fundamentally altering the structure of work by automating routine tasks while simultaneously creating new roles and responsibilities. This dual effect is central to understanding its impact on employment. Rather than eliminating jobs outright, AI redistributes tasks across occupations, often shifting demand toward roles that require higher cognitive, technical, or interpersonal skills.

This process, described as task reallocation, produces uneven outcomes across the workforce. Workers with access to training and digital skills are more likely to benefit from AI adoption, moving into augmented roles where human capabilities complement machine intelligence. In contrast, workers in routine or low-skill roles face a higher risk of displacement, particularly when reskilling opportunities are limited.

The research highlights that productivity gains driven by AI do not automatically translate into inclusive growth. Organizations may achieve efficiency improvements through automation, but without deliberate strategies for workforce transition, these gains can widen existing inequalities. The emergence of generative AI further accelerates this dynamic, introducing new job categories while simultaneously disrupting established roles.

Employment outcomes are shaped by organizational and institutional conditions. Companies that invest in job redesign, training, and human–AI collaboration are more likely to create inclusive environments. In contrast, those that prioritize cost reduction through automation without supporting workforce adaptation risk reinforcing exclusion.

AI is not simply a technological shift but a structural transformation of labor markets. Its effects depend on how tasks are redistributed and whether workers are equipped to adapt to new demands.

Algorithmic management is redefining job quality and worker experience

The study finds that AI is reshaping the nature of work itself. Algorithmic management systems are increasingly used to monitor performance, allocate tasks, and evaluate employees, introducing new forms of control and oversight in the workplace.

These systems offer clear benefits in terms of efficiency and coordination. By analyzing large volumes of data, AI can optimize workflows and support decision-making processes. However, the research shows that these advantages are accompanied by significant risks to job quality.

• Intensification of surveillance: Data-driven monitoring systems can track employee behavior in real time, raising questions about privacy and autonomy. Workers may experience reduced control over their tasks and increased pressure to meet algorithmically defined performance metrics.
• Fairness and transparency. When AI systems are used in performance evaluation or recruitment, their decision-making processes are often opaque. Employees may find it difficult to understand how decisions are made or to challenge outcomes they perceive as unfair. This lack of contestability undermines trust and can lead to perceptions of injustice.
• Legal and regulatory gaps: In many cases, it is unclear who is accountable for decisions made by AI systems. This ambiguity creates challenges for enforcing worker rights and addressing grievances, particularly in cases of discrimination.

The research underscores that job quality in AI-driven workplaces is shaped not only by technology but by governance. Without clear safeguards, the same systems that enhance efficiency can erode autonomy, dignity, and fairness.

Inclusion gains coexist with risks of exclusion and discrimination

AI is both an enabler of inclusion and a driver of exclusion. On one hand, AI technologies can expand access to employment by reducing barriers and supporting diverse work arrangements. Assistive technologies, adaptive interfaces, and AI-enabled accommodations have improved opportunities for workers with disabilities and other marginalized groups.

AI can also enhance recruitment processes by standardizing evaluation criteria, potentially reducing certain forms of human bias. When properly designed and audited, algorithmic systems can improve consistency and broaden access to opportunities.

However, the study makes clear that these benefits are conditional. AI systems can reproduce and amplify existing biases if they are trained on historical data that reflects discriminatory patterns. Even when explicit demographic variables are removed, proxy variables can encode similar biases, leading to unequal outcomes.

The risk of digital exclusion is another major concern. Workers without access to digital skills or infrastructure may be excluded from AI-mediated labor markets altogether. This creates a divide between those who can engage with new technologies and those who cannot, reinforcing existing social and economic inequalities.

The study also identifies the emergence of dual labor markets, where highly skilled workers benefit from AI-driven productivity gains while others are pushed into precarious or low-quality jobs. This polarization reflects broader structural changes in the economy, where technological advancement does not benefit all groups equally.

These findings highlight that inclusion is not an automatic outcome of technological progress. It requires deliberate efforts to address structural barriers and ensure that AI systems are designed and deployed with equity in mind.

Governance emerges as the decisive factor shaping AI outcomes

The impact of AI on labor markets ultimately depends on governance. Technological innovation alone cannot ensure inclusive outcomes. Instead, a coordinated approach involving organizations, policymakers, and workers is needed to manage the transition effectively.

To address this challenge, the authors propose a hybrid governance model that integrates four key dimensions: technical oversight, organizational practices, workforce development, and regulatory frameworks.

• At the technical level, transparency and auditing are essential for identifying and mitigating bias in AI systems. Continuous monitoring and clear documentation can help ensure that algorithms operate fairly and accountably.
• Organizational practices: Responsible human resource management, including meaningful human oversight in decision-making processes, can reduce the risks associated with automated systems. Ensuring that employees have the ability to question and appeal decisions is key to maintaining trust.
• Workforce development: Investing in upskilling and reskilling programs can help workers adapt to changing job requirements, reducing the risk of displacement and enabling more inclusive participation in AI-driven economies.
• Regulatory frameworks: Participatory approaches that involve workers and other stakeholders in decision-making processes can help ensure that policies reflect diverse perspectives and protect fundamental rights.

These governance dimensions are interdependent. Weakness in one area can undermine progress in others, highlighting the need for coordinated strategies that address multiple aspects of AI adoption simultaneously.