Purpose: Deepfakes, which first appeared in 2017, use artificial intelligence (AI) to generate modified but incredibly realistic digital content, including political satire, posing serious moral, legal, and societal concerns. The purpose of the study is to identify the evolving landscape of deepfake technology, with a focus on its psychological and societal impacts.

Design/methodology/approach: The researchers carried out a comprehensive bibliometric analysis of research from 2019 to 2024, using data obtained from the Web of Science and Scopus bibliographic databases. After combining 688 bibliographic entries from both databases using Bibliometrix R. After eliminating nonEnglish publications and studies irrelevant to the bibliometric scope of deepfake research, a total of 463 articles were included for further study.

Findings: Based on our results, the researchers found that academic scholarships on detection techniques, misinformation, and social impact has grown rapidly, particularly in China, the United States, and India. Six main study areas have emerged, such as deepfake detection methods, information integrity, machine learning, social media influences, forensic analysis, and facial manipulation. The findings reveal that the increasing complexity of deepfake generation and its consequences for ethical concerns significantly affects digital trust, privacy infringement, and the spread of misinformation.

Originality/value: More interdisciplinary techniques combining AI, social sciences and ethics are required despite notable gains in detection. To reduce the risks associated with deepfakes, the study highlights the significance of strong detection technologies, regulatory frameworks and public awareness. The psychological effects of deepfake exposure and the creation of moral standards for appropriate AI use should be the focus of future studies.