Institutional Repository of Vilnius University
Dirbtinis intelektas kaip įrankis sukčiavimui: balso sintezavimo atvejis /
Abstract
dc:descriptionThe aim of this research is to identify factors that can mitigate the frequency of fraudulent activities, in which criminals use speech synthesis technology. This study presents the Routine Activity Theory, exploring its application in investigating cyber-related crimes. Additionally, it discusses speech synthesis technology based on machine learning algorithms, detailing two primary methods—text-to-speech (TTS) and speech-to-speech (STS)—and their operational principles. To determine the factors that can prevent such fraudulent activities, two qualitative studies were conducted. The qualitative analysis of 9 speech synthesis programs revealed limitations such as mandatory registration, paid voice cloning services, and short durations of synthesized speech recordings. These limitations act as „safeguards“, potentially deterring criminals from committing fraud. Conversely, the analysis also highlighted the advantages of these programs, such as extensive libraries of synthesized voices and real-time voice synthesis capabilities, which might contribute to an increase in fraud. Furthermore, a qualitative content analysis of news articles presented 7 different cases of fraud, described in a total of 38 articles. This analysis identified three critical factors that can reduce the risk of falling victim to speech synthesis fraud: lack of required funds, verification of a person's identity by checking their phone number, and directly calling the person, who's being impersonated by the fraudster.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Jankauskas, Marius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210648456&prefLang=en_US