Artificial intelligence (AI) has shifted from peripheral utilities to core newsroom infrastructure, assisting tasks from transcription and translation to drafting, packaging, personalisation, moderation, and analytics. This article evaluates how AI augmentation is transforming labour and skill profiles in contemporary newsrooms. Through an integrative review spanning journalism study, human–computer interaction (HCI), computer supported cooperative work (CSCW), labour economics, media law/ethics, and organisational change, it maps role recompositing across editorial, data/visual, audience, product, and engineering teams. We synthesise evidence on productivity, quality, trust, and job satisfaction; identify risks including deskilling, opacity, bias, covert automation, safety/confidentiality threats, and the erosion of entry level pathways; and propose a layered human in the loop pipeline with governance mechanisms (disclosure, review thresholds, audit trails, and provenance controls). The analysis advances a competency framework that combines beat expertise, verification craft, model literacy, workflow design, and ethical/legal fluency, and outlines policies and training pathways to align augmentation with public interest outcomes. We conclude that sustaining journalistic quality and decent work requires deliberate sociotechnical design, investment in complementary skills, and incentive realignment away from pure engagement optimisation toward metrics that value accuracy, diversity, and civic impact.
Adedowole et al. (Wed,) studied this question.