Show simple item record

A FRAMEWORK ANALYSIS OF DEEPFAKES: USING SWOT AND FMEA TO CALCULATE THE RISK POSED BY DEEPFAKES

dc.contributor.advisorFisher, Douglas
dc.contributor.advisorWisniewski, Pamela
dc.creatorKoul, Kastur
dc.date.accessioned2023-05-17T20:51:16Z
dc.date.available2023-05-17T20:51:16Z
dc.date.created2023-05
dc.date.issued2023-03-27
dc.date.submittedMay 2023
dc.identifier.urihttp://hdl.handle.net/1803/18231
dc.description.abstractWhile it is advisable to be cautious in the face of new technological developments, aggressive negativity towards a piece of technology can cause fear and distrust in the technology. This is the case for deepfakes, which have received negative press since their first appearance in media in 2017. Published media articles often state the potential risks of deepfakes in ways that make the technology seem like it can only be used for negative purposes without analyzing the risks using risk analysis methods. In this thesis, I use the SWOT analysis technique to list the strengths, weaknesses, opportunities, and threats of deepfakes. I then analyze the uses of deepfakes presented by the opportunities and threats using the FMEA method to score the severity of the consequences, probability of use, and likelihood of detection. These scores are then used to calculate the expected risk of the uses of deepfakes. Results show that deepfaked pornography poses the most severe risk, and the uses of deepfakes in entertainment the least risk. Overall, there is a nonnegligible risk posed by deepfakes. This analysis should open the conversation of deepfakes to more analytical and cautious rather than overtly negative views.
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.subjectdeepfakes
dc.subjectethics
dc.subjectrisk analysis
dc.subjectSWOT
dc.subjectFMEA
dc.subjectGAN
dc.subjectencoder/decoder networks
dc.subjectCNN
dc.subjectLSTM
dc.titleA FRAMEWORK ANALYSIS OF DEEPFAKES: USING SWOT AND FMEA TO CALCULATE THE RISK POSED BY DEEPFAKES
dc.typeThesis
dc.date.updated2023-05-17T20:51:16Z
dc.type.materialtext
thesis.degree.nameMS
thesis.degree.levelMasters
thesis.degree.disciplineComputer Science
thesis.degree.grantorVanderbilt University Graduate School
dc.creator.orcid0009-0007-5598-9505


Files in this item

Icon

This item appears in the following Collection(s)

Show simple item record