This research explores the feasibility and limitations of using generative artificial intelligence (AI) as a collaborator in systematic literature reviews. Focusing on ethics in AI startups as an illustrative domain, the study compares human and AI-assisted screening of academic publications to assess consistency, efficiency, and interpretive differences. The findings highlight that while AI tools can enhance transparency and reduce manual workload, they also tend to apply rigid criteria that may overlook contextual nuances captured by human evaluators. The research contributes to ongoing debates about the role of generative AI in academic knowledge production and underscores the need for protocols that combine machine precision with human judgment.
Hybrid Approaches to Systematic Literature Reviews on Generative AI Ethics: Combining Human and AI Judgement
Marco Di Giovanni
Primo
;Filomena Pagnozzi;Gilda Antonelli
2025-01-01
Abstract
This research explores the feasibility and limitations of using generative artificial intelligence (AI) as a collaborator in systematic literature reviews. Focusing on ethics in AI startups as an illustrative domain, the study compares human and AI-assisted screening of academic publications to assess consistency, efficiency, and interpretive differences. The findings highlight that while AI tools can enhance transparency and reduce manual workload, they also tend to apply rigid criteria that may overlook contextual nuances captured by human evaluators. The research contributes to ongoing debates about the role of generative AI in academic knowledge production and underscores the need for protocols that combine machine precision with human judgment.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
