Deepfake Threats: 2024–2025 Key Facts
Explosive Growth
Global deepfake incidents rose 10x in the past year.
3000% increase in deepfake-based fraud since 2022.
Financial scams using deepfakes surged 700% last year.
Main Targets & Risks
88% of deepfake scams target cryptocurrency.
96% of deepfake videos are non-consensual pornography.
400+ businesses a day targeted by deepfake CEO/video scams (USA).
Average business loss per deepfake scam: $500,000.
Notable Incidents
$25 million stolen: AI Zoom call scam (Hong Kong).
$620,000 stolen: Face-swapped video call scam (China).
Pentagon explosion deepfake: Sparked panic and moved markets (USA).
US elections: AI-Biden robocalls tried to disrupt voting.
AI Voice Cloning
Only 3 seconds of audio needed to fake a voice convincingly.
1 in 10 people received a scam call using an AI-cloned voice.
77% of AI voice scam victims lost money.
Detection is Hard
Humans spot AI fakes correctly only 51% of the time.
Top AI detectors catch only 65% of advanced deepfakes.
Deepfake attacks on ID verification checks up 704% last year.
Nepal Snapshot
68% of Nepali youth know about generative AI; 46% use it.
635+ child cybercrime cases (2023–24); 260% rise, mostly AI-edited images.
No current law in Nepal specifically bans deepfakes.
Dangerous Trends & What’s Next
$40 billion: Predicted global fraud losses from AI by 2027.
By 2026, 30% of companies may find standard ID checks obsolete.
Real-time deepfake impersonation is here.
Public trust in audio and video evidence is falling fast.