A comprehensive analysis of voice deepfake incidents in Malaysia and their implications for digital trust
Data Analysis
Security Impact
Global Context
UiTM Tapah
Course: ICT652
Group: A4CDCS2306A
2024745815
2024974573
2024741073
2024916761
Live demonstration of voice cloning technology using our team's voice samples
Original vs Cloned Voice
Original vs Cloned Voice
Original vs Cloned Voice
Original vs Cloned Voice
Can you distinguish between the original and cloned voices? This demonstrates the sophistication of modern voice cloning technology.
Human Perception
Increasingly difficult to detect
Current Detection
Limited effectiveness
AI Arms Race
Technology vs Detection
Investigate and analyze high-profile voice deepfake incidents in Malaysia to understand their impact on digital trust and security
Document and categorize all voice deepfake incidents in Malaysia
Examine patterns, methods, and targets of voice deepfake attacks
Evaluate the effectiveness of current detection and prevention measures
Propose strategies to combat voice deepfake threats
Voice deepfakes represent one of the most significant threats to digital trust in the modern era
Voice authentication bypass, financial fraud, identity theft
Evidence manipulation, defamation, jurisdictional challenges
Erosion of trust, misinformation, public safety concerns
Malaysia's Unique Position
Rapid digital adoption + Multilingual society = Complex challenge landscape
Deepfake audio of political figures making controversial statements circulated during election period
Criminals used voice deepfakes to bypass bank voice authentication systems, resulting in significant financial losses
Voice deepfake used to impersonate executives for obtaining sensitive business information
Fake endorsements and investment schemes using celebrity voice clones
Understanding the tools behind voice deepfake technology
Commercial platforms with proprietary technology
Community-driven, freely available tools
Premium voice synthesis with emotional intelligence
Leading voice AI with sub-second latency, instant voice cloning
Real-time voice cloning and synthesis
AI voice actors for games and media
Core piece of global voice intelligence, outstanding speech technology
Enterprise-grade voice synthesis platform
Open-weight TTS model with 82 million parameters.
A Breakthrough in Emotionally Expressive and Duration-Controlled Auto-Regressive Zero-Shot Text-to-Speech
A Frontier Open-Source Text-to-Speech Model
State-of-the-art speech model for expressive voice generation
Leading Open Source Voice Cloning AI Model
A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching
The trust paradox: As AI voice technology improves, the harder it becomes to distinguish real from fake
The rise of voice deepfake technology presents both unprecedented challenges and opportunities for Malaysia's digital future.
Stay alert to emerging threats
Evolving security measures
Multi-stakeholder approach
The future of digital trust depends on our actions today.
Thank You
Questions & Discussion
Presented by: Amir Hafizi, Muhammad 'Adli, Nik Muhammad Haziq, Mohamad Bukhari