The Broken Ladder: Generative AI, Digital Feudalism, and the Structural Collapse of Entry-Level Tech Employment for Global South Graduates

Lokesh Kumar · Zenodo (CERN European Organization for Nuclear Research) · 4 Nov 2026

Where this sits in the debate

Will AI destroy more jobs than it creates? · its finding: it depends (why studies differ)
Will AI widen inequality? · its finding supports the claim (why studies differ)

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

Abstract

The rapid diffusion of generative AI coding assistants (e.g., GitHub Copilot, Cursor, Claude Code) has dismantled the foundational 'escalator' of software engineering employment in the Global South: entry-level offshore code maintenance, boilerplate web scripting, and manual quality assurance. While advanced economies capture productivity premiums and capital windfalls, peripheral developing nations face an asymmetric structural crisis. This whitepaper analyzes the emergence of 'Digital Feudalism'—an economic regime where Global South software graduates are displaced from traditional IT services arbitrage and relegated to uncompensated prompt telemetry generators for centralized Northern foundation models. Drawing on labor telemetry across Pakistani tech corridors and a multi-institutional survey of computer science graduates from non-metropolitan public universities (2024–2026), we demonstrate that junior hiring postings declined by 42.8% while licensing costs for frontier AI tools consume up to 85% of regional entry-level developer salaries. We articulate a four-pillar sovereign counter-framework to restore socioeconomic mobility for peripheral engineers: Sovereign Vernacular Data Engineering: Transitioning from commoditized web development to high-value indigenous corpus creation. Architectural Verification Curricula: Pivoting computer science education from rote syntax generation to formal verification, distributed resilience, and AI hallucination audit. Federated Compute Cooperatives: Establishing subsidized, national cloud-GPU access for peripheral students. Open-Access Civic Tooling: Anchoring developer productivity in tangible local problems (such as agro-informatics and environmental telemetry). Open-Access Research Systems: • System Portal: https://agri-idas.tech/ • Vernacular NLP Corpus: https://huggingface.co/datasets/lokeshsonara/dhatki-agri-intents

About this publication

Preprint. Posted on a preprint server or repository (arXiv, SSRN, Zenodo and others) before or without peer review. Anyone can post there.

How sources are ranked.

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

0 citations recorded by OpenAlex — too few 2026 papers in this collection for a fair comparison.

Reaches a different conclusion

Papers on the same question whose finding points the other way, as read from the abstracts by an AI model (AI-written). Why studies differ.

Forecasting the Economic Effects of AI
Jobs: its finding supports the claim · Ezra Karger et al. · 2026
The Future of Jobs Report 2025
Jobs: its finding goes against the claim · World Economic Forum · 2025
Experimental evidence on the productivity effects of generative artificial intelligence
Inequality: its finding goes against the claim · Shakked Noy and Whitney Zhang · 2023
The ABCs of Who Benefits from Working with AI: Ability, Beliefs, and Calibration
Inequality: its finding goes against the claim · Andrew Caplin et al. · 2025

The Broken Ladder: Generative AI, Digital Feudalism, and the Structural Collapse of Entry-Level Tech Employment for Global South Graduates

Abstract

About this publication

Reaches a different conclusion

Related papers