Artificial Intelligence (AI) and the Future of the Engineering and Computing Workforce: A Systematic Review of Gray Literature and Document Analysis of U.S. Reports (2020–2025)

Synthesis: Fletcher and colleagues (2026) conduct a systematic review of U.S. gray literature and document analysis of 22 reports (2020–2025) to understand how AI is reshaping the engineering and computing workforce and what this means for higher education. Framing the tension between rapid technological change and urgent policy decisions as the "Dual Train Problem," the review synthesizes projections from sources such as the Bureau of Labor Statistics, MIT's Shaping Work Project (1.6–3.2 million U.S. jobs displaced over two decades; up to 30% automatable by the mid-2030s), and the World Economic Forum. The findings imply that while some engineering and computing fields will grow, AI will replace many tasks (reducing jobs and working hours) even as new AI-oriented roles emerge, leading the authors to recommend that higher education prioritize durable AI competencies, Ethics and AI Governance, and skill-based credentials aligned with emerging roles (e.g., Prompt Engineering, AI auditing, AI policy) to sustain human-centered engineering.

The Dual Train Problem

The paper opens with the "Dual Train Problem": the rapid advancement of AI and automation runs in tension with the urgency of policy and institutional decisions, and higher education must adapt proactively. The authors note that AI is now believed to affect white-collar jobs — not only the middle- and low-skilled occupations that prior automation targeted — and that attitudes toward AI's impact diverge sharply between those who foresee productivity and new roles (data scientists, AI programmers) and those who perceive displacement, inequality, and polarization.

Workforce projections and scenarios

Drawing on economic and labor-market forecasts, the review surfaces a range of projections:

• Bureau of Labor Statistics (2023–2033) projections now account for AI exposure across several computer and engineering occupations, acknowledging heightened automation risk.
• MIT's Shaping Work Project estimates 1.6–3.2 million U.S. jobs could be displaced by AI-driven automation over two decades; more ambitious projections suggest up to 30% of jobs may be automatable by the mid-2030s.
• Four scenarios for 2030 (after Ellis 2025): supercharged progress (exponential AI advance outpacing governance), age of displacement (automation outpacing reskilling, generating social fractures), co-pilot economy (gradual AI-augmented work through human–AI collaboration), and stalled progress (uneven skill diffusion widening adoption gaps).
• Microsoft (2025) finds skills in analytical thinking, resilience, Ethics, and digital literacy are twice as likely to be required in job postings.

The review's central conclusion from these projections is that engineering and computing face a transformation — some growth in selected fields, but AI replacing tasks and reducing jobs and working hours, alongside the emergence of new AI-oriented roles requiring generative AI upskilling.

What this means for practice

• Instructors. Prioritize durable AI competencies that transfer across evolving roles rather than tool-specific skills, since the reviewed projections show task replacement alongside the emergence of new AI-oriented roles.
• Instructors. Embed Ethics and AI Governance in engineering and computing formation as required content, not add-ons, as the authors recommend for faculty and deans.
• Administrators. Adopt skill-based credentials aligned with emerging roles such as prompt engineering, AI auditing, and AI policy.
• Administrators. Coordinate curriculum, industry partnerships, and resource allocation deliberately against the "Dual Train Problem," because the review's central conclusion is that the institutional train is moving slower than the technological one and that human-centered engineering depends on that coordination.
• Administrators. Weight durable human skills in program outcomes — Critical Thinking and digital literacy were reported as twice as likely to be required in job postings (Microsoft, 2025).

Limitations

• The corpus is 22 reports, narrowed from 2,320 Google Search results through PRISMA-style screening; searches ran through Google rather than a bibliographic database because the target was gray literature.
• Scope is deliberately narrow: only U.S.-based sources published 2020–2025 that contained specific projections for engineering and computer science fields, so the authors caution that findings should be interpreted as representative of those two fields only.
• The authors state that the recommendations and curricular interventions proposed for institutions have not yet been empirically tested and call for longitudinal work on their feasibility and effectiveness.
• The analysis is bounded by the scope of the existing reports and their forecasts, which may not project the full extent of AI-driven change in engineering and computing roles.

Connections to the knowledge base

The paper extends the knowledge base's coverage of AI and work, complementing empirical frameworks like the Workforce Readiness Level and connecting to Workplace Learning, AI Literacy, and AI Governance concepts. It frames Curriculum Design for engineering and computing as a strategic response to labor-market transformation, and positions Higher Education institutions as key actors in workforce preparation.

Citation

Fletcher, T. L., Webb, M. E., Alharbi, A., & Fletcher, T. (2026). Artificial Intelligence (AI) and the Future of the Engineering and Computing Workforce: A Systematic Review of Gray Literature and Document Analysis of U.S. Reports (2020–2025). ASEE Annual Conference & Exposition, Paper ID #53492.