Truth or Fiction The Art of Unveiling DeepFakes in Text, Audio, Image
summary
A rundown of practical signals and tools: a CEO voice deepfake almost triggers a bank transfer, image frequency analysis and EXIF metadata checks catch many fakes, and spectrogram/pitch analysis plus excitation patterns flag synthetic audio. Detectors like ZeroGPT/GP Z and GLTR top benchmarks but top accuracy sits near 76% detection ceiling, while the open RAID 10M-document benchmark lets teams compare models; watermarking remains the most promising prophylactic.
AI-written summary · about
chapters (8)
- 0:00🚀 Introduction: Deepfake detection and the challenges ahead
- 4:00What are deepfakes? The dangers of AI-generated fake content
- 8:00The importance of critical thinking and triangulating sources
- 12:00Detecting deepfake text: Tools and strategies
- 16:00Analyzing AI-generated images and how to spot them
- 20:00Audio detection: How spectrograms and pitch analysis help
- 24:00Video deepfakes and tools to detect them
- 26:00Summary: The importance of being critical of information
related talks
30mReality Under Attack: Just How Serious Is the Deep Fake Threat?
26mDeepfakes, audiofakes, and the future of media
30mAI-Generated Synthetic Media, aka Deepfakes and its impact
41mDeepfake Threats in Business: What You Need to Know
47mFighting AI with AI
41mThe Deepfake Supply Chain: Weaponizing Synthetic Media into Modern Cybercrime Operation
46mTackling Deepfakes, Wars, and Other Security Threats in the GenAI World
27mIs now the time to solve the deepfake threat?