Abstract
Generative artificial intelligence is usually discussed in computing education as a challenge to teaching, assessment, and academic integrity. I argue that an equally consequential issue is emerging within computing education research (CER): federal funding disruption is weakening the graduate research pipeline just as AI is reducing the visible labor required to generate software, analyses, synthetic data, documentation, proposals, and manuscripts. These forces should not be conflated. Funding retrenchment is the immediate threat to projects and trainee support; AI is a contingent amplifier that may strengthen well-funded teams or allow faculty to maintain scholarly output with fewer students and less grant seeking. CER is especially vulnerable because much of its consequential work is relationship-intensive, longitudinal, and context-dependent. It is also upstream infrastructure for computing workforce development: CER-trained researchers determine whether AI-mediated instruction produces durable learning, how curricula and assessment should change, how teachers adapt, and who gains access to computing careers. I develop a prospective, testable account of how funding loss, AI-enabled productivity, and growth in self-financed professional master's education may separate publication and enrollment from researcher formation. I then propose a CER capacity dashboard and a revenue-preserving MS research-training compact designed to protect doctoral formation, authentic field research, methodological diversity, equitable entry into research careers, and the expertise needed to prepare an adaptable computing workforce.