The Data Diva E183 - Debesh Choudhury, Ph.D. and Debbie Reynolds

"The Data Diva" Talks Privacy Podcast

Send us a text

Debbie Reynolds “The Data Diva” talks to Debesh Choudhury, PhD, Information Security Researcher (India). We discuss the application of image recognition for security, with both expressing concerns about the accuracy and sensitivity of the technique, particularly in matching images and creating databases of individuals. We also highlight the potential vulnerabilities in image security, emphasizing the need for complex hashes and the development of quantum-proof techniques. Additionally, the conversation touches on the limitations of voice recognition as a biometric factor for authentication, focusing on the ease of spoofing and the necessity of multiple factors in security measures. Debbie expresses deep concern about the rapid advancement and potential dangers of deepfake technology, citing a news story about a significant financial loss resulting from a deepfake video call. Debesh emphasizes the need for research to detect deepfakes and discusses the multifaceted impact of deepfake technology, including its potential positive applications in entertainment and the movie industry, as well as its negative implications in political scenarios and personal lives. The discussion also touches on the pressure for companies to adopt AI and the potential risks associated with overreliance on artificial intelligence, drawing parallels to historical industry shifts such as the adoption of Linux by Microsoft. Debesh raises concerns about the widespread adoption of biometrics and its potential threats to privacy, citing examples of government orders and the impact of the pandemic. We discuss the negative impact of such technologies on individuals and advocated for their rights. We discuss the need for logical thinking and education in the context of introducing AI and robotics courses in high schools, emphasizing the importance of nurturing critical thinking skills in students and his wish for the future of Data Privacy.

Support the show