The Threat of AI Automation on Employment Termination: A Systematic Literature Review

Ilham Prasetya Wibowo, Cucu Hayati, Tri Wahjoedi · Global Economic Social and Development Review · 10 Oct 2026

Where this sits in the debate

Will AI destroy more jobs than it creates?

Direction of the finding as read from the abstract by an AI model (AI-written).

Key takeaway AI-written

A PRISMA systematic review of 25 peer-reviewed articles (2016-2026) on AI-driven job terminations. It identifies four displacement patterns, including hierarchical task substitution and opaque algorithmic layoffs, finds service, IT and labor-intensive sectors and middle-skilled workers most exposed, and proposes reskilling, algorithmic accountability rules and union engagement as mitigations.

The paper in three steps AI-written

Written by an AI agent from the paper's abstract, which is the primary source and is quoted below.

The question

How does AI automation lead to job terminations, which workers are most exposed, and what can limit the harm?

The approach

A systematic review following PRISMA 2020: 25 peer-reviewed articles (2016-2026) selected from 520 Google Scholar records, analysed by thematic synthesis and keyword co-occurrence mapping.

The findings

The review identifies four displacement patterns: task substitution up the hierarchy, skill degradation and polarisation, acceleration with generative AI, and opaque algorithmic layoffs. Service, IT and labour-intensive sectors and middle-skilled workers face the most pressure.

How to read it: review. A synthesis of existing studies — breadth over novelty; quality depends on what it includes and excludes.

Limitations and counterpoints AI-written

Only 25 studies from one search engine, and only one study on Indonesia, so the authors present the findings as context rather than direct evidence for developing countries.

Abstract

The rapid diffusion of artificial intelligence (AI) across industries has intensified concerns regarding labor substitution, yet scholarly literature on AI-driven employment termination remains fragmented. This study systematically identifies displacement mechanisms, maps vulnerable sectors and workforce categories, and synthesizes mitigation strategies. Following the PRISMA 2020 protocol, a Systematic Literature Review (SLR) was conducted via Google Scholar across three search sessions, yielding 25 eligible peer-reviewed articles (2016–2026) from 520 initial records. Data were analyzed using thematic synthesis complemented by keyword co-occurrence bibliometric mapping in VOSviewer. Results indicate four interrelated displacement patterns: hierarchical task substitution, skill degradation and labor market polarization, acceleration coinciding with generative AI, and opaque algorithmic layoff practices. Service, information technology, and labor-intensive sectors are identified as most vulnerable, with middle-skilled workers experiencing the greatest structural pressure. Effective mitigation requires a simultaneous three-pillar approach: competency-based reskilling, algorithmic accountability regulation, and collaborative labor union engagement, with proactive timing highlighted as critical. These findings offer a synthesized evidence base for policymakers and organizations navigating an AI-augmented labor market. Given that only one reviewed study directly examined Indonesia, these findings serve as contextual implications rather than direct empirical assertions, underscoring the necessity for future primary research in developing national contexts.

About this publication

Academic journal. Published in an academic journal — typically peer-reviewed before publication, though indexing alone doesn't guarantee review depth.

Published: 2026 (recently)
Citations: 0 (OpenAlex)