Unsupervised Learning podcast

A Conversation with Shiladitya Sircar from BlackBerry on DeepFake Threats

0:00
44:44
Manda indietro di 15 secondi
Manda avanti di 15 secondi

In this conversation, I speak with Shiladitya Sircar, Senior VP of Product Engineering and Data Science at BlackBerry.

We talk about:

The Rise of Deepfakes and Cyber Threats
Innovation Meets Malicious Intent: Deepfakes are not just a tech novelty; they’re a growing threat. From text-based phishing to hyper-realistic fake videos and audio, the landscape of cyber threats is evolving rapidly. Deepfake technology can clone voices, making it easier for cybercriminals to impersonate individuals and bypass security measures.

Understanding Identity Compromise
Voice Cloning Dangers: Our brains are wired to trust familiar voices, making voice cloning particularly insidious. We share a chilling story about a cybercriminal impersonating Ferrari’s CEO. The attacker’s deepfake was so convincing that it almost led to a major scam.

The Impact on Trust
Eroding Trust in Systems: Deepfakes can undermine trust in institutions and systems, much like traditional scams but with a high-tech twist. Beyond individual attacks, deepfakes can manipulate public opinion and even influence elections. Organizations need to train employees to spot deepfakes, and there’s a pressing need for laws that specifically address deepfakes and identity spoofing.

And more

Intro (00:00:00)
Main Cyber Threats from Deepfakes (00:00:56)
Identity Compromise Explained (00:02:47)
Impact of Deepfakes on Trust (00:06:23)
Deepfakes in Attack Chains (00:08:15)
Case Studies of Deepfake Attacks (00:09:41)
Emerging Threat Landscape (00:13:56)
Defending Against Deepfake Attacks (00:15:07)
Regulatory Frameworks Needed (00:16:28)
The Role of Education and Technology (00:18:57)
Future of Content Authenticity (00:20:53)
Legislation and Authenticity Mechanisms (00:22:04)
Real-Time Deepfake Validation (00:23:18)
Government and Industry Partnership (00:24:07)
Media Forensic Research (00:24:23)
Zero Knowledge Proofs (00:25:36)
Content Provenance and Authenticity (00:26:52)
Trust Network Expansion (00:28:00)
Puppeteering Technology (00:29:20)
Stream Authentication Challenges (00:30:21)
Hardware-Level Trust (00:32:00)
Fragmentation in Standards (00:32:29)
Trust in Communication Protocols (00:33:51)
Collaboration for Solutions (00:35:22)
Apple's Unique Position (00:36:47)
Erosion of Trust (00:37:31)
AI Agents for Detection (00:38:11)
Short-term and Long-term Solutions (00:38:45)
Awareness and Education (00:41:23)
Predictions for Deepfake Technology (00:41:48)
Community Action Against Deepfakes (00:43:09)
Learning More About BlackBerry's Work (00:43:29)

Become a Member: https://danielmiessler.com/upgrade

See omnystudio.com/listener for privacy information.

Altri episodi di "Unsupervised Learning"