Keyless Webinar Recap: The Deepfake Dilemma

In our latest webinar, hosted by Darius Montvila, Sales Director North America and Paolo Gasti, Associate Professor at NYIT and our CTO, explored the growing challenge of deepfakes and their implications for identity verification in the digital age. The session delved into the evolution of deepfakes, the threats they pose, and innovative solutions to counter these threats.

The AI Evolution and the Beginnings of Deepfakes

The webinar kicked off with a historical overview of AI and deepfakes. Darius highlighted that the exploration of AI for image manipulation dates back to the 1990s, with a significant milestone in 2014 when Ian Goodfellow introduced Generative Adversarial Networks (GANs). Paolo provided insights into how GANs work, explaining that they involve two neural networks – a generator and a discriminator – that work in tandem to create increasingly realistic images.

Emergence and Risks of Deepfakes

In 2017, the term "deepfake" was coined on a subreddit dedicated to videos using face-swapping technology. Today, deepfakes have evolved from crude manipulations to hyper-realistic content, raising concerns about extortion, blackmail, misinformation, and political subversion. The dual-edged nature of deepfakes was emphasized, with their potential for both creative and malicious applications.

High-Profile Incidents

Darius discussed several high-profile deepfake incidents, including a fake Biden robocall urging New Hampshire Democrats to stay home, a Hong Kong company worker losing £20 million in a deepfake video call scam, and the CEO of WPP, the world's largest advertising firm, being targeted. These examples underscore the growing sophistication of deepfake attacks and their potential impact.

Threats to Identity Verification

The webinar then shifted to the specific threats deepfakes pose to identity verification. Paolo explained presentation attacks and injection attacks, where fake images or videos are used to spoof identity verification processes. One notable example mentioned was using a deepfake of Elon Musk to bypass security checks.

Role of Liveness Detection

Liveness detection emerged as a critical defense against deepfakes. Paolo differentiated between traditional detection methods and the more advanced passive liveness detection used by Keyless. Passive liveness detection identifies subtle indicators such as skin color variations, textural patterns, light reflections, and unique traits in the periocular region. These indicators help distinguish between live individuals and static images or videos.

Keyless employs a multi-layered defense strategy against deepfake threats, as recommended by Gartner. Darius emphasized Keyless' commitment to privacy-preserving, multi-factor authentication, leveraging passive liveness detection. This approach ensures robust authentication while maintaining user privacy.

Wrap-Up

The aim of the Deepfake Dilemma webinar was to provide valuable insights into the origins, risks, and countermeasures associated with deepfakes. If you didn't get a chance to watch the webinar live, you can access the webinar recording here.