PRIVACY PROTECTION: AI VOICE BIOMETRICS AND EMOTION RECOGNITION

Sammanfattning: The increased awareness of data collection has in recent years raised the debates on privacy and ethical questions regarding personal data-collection. Companies want to make use of voice biometrics technology, which causes ethical and privacy related challenges. This report investigates: the existing regulations, standards, and guidelines; the technical aspects of voice biometrics; and the user’s acceptance of the technology. The objective is to use the research data for the purpose of improving the existing privacy requirements to mitigate the ethical and privacy related risks. The improvements will address the concerns of ethicists and users and contribute to bridging the gap and build trust between them and the companies using the technology. To collect the data, literature review and a survey is conducted. Results from the survey sent to university students indicate a negative outlook on the use of voice biometrics as a form of data collection. Results show that the willingness to share more data in exchange for a personal experience is low. Furthermore, there is also a lack of trust towards companies when it comes to data-protection. The results show that companies use emotion recognition for the removal of manual user input to enhance the user experience. We identified the technical details of what data is extracted from the voice as well as the process of voice biometrics emotion recognition. The results show ethical issues with emotion recognition and voice biometrics. Existing laws and regulations are broken down to requirements and by analyzing the combination of the extracted data, we provide the abovementioned privacy requirements for bridging the gap between the stakeholders.