The higher output quality of neural machine translation (NMT) has fostered research into the possibility of using it for literary texts, leaving the translator the finishing touch through post-editing. However, as previous studies have shown (Moorkens et al.,, 2018; Guerberof-Arenas and Toral, 2020, 2022), the creativity that characterises literary translators’ reading and writing is hindered by the predominantly literal MT output. While subscribing to the idea of leaving the human translator in charge of literary texts, our article tests another use of technology—namely aligners and NMT to assist in the revision of previously published translations which show good quality and relevance to today’s world, yet contain factual errors, omissions, misinterpretations and dated lexical or syntactic choices. Two chapters from Sinclair Lewis’ The Job (1917) and Rachel Carson’s Silent Spring (1962) were aligned with existing translations (Italian for the former, published in 1955 and 1970, and Finnish for the latter, published in 1963) and with the output produced by DeepL Pro to see whether some of its segments could replace the parts to be revised. Then, professionals working in literary translation were asked if 1) they were in favour of the use of technologies in the revision process, and 2) if the method we proposed was viable.