AI has become a force multiplier for scammers. It enables them to accomplish in seconds what once took hours—at massive scale and with startling accuracy. Fraudsters have moved from a cottage-industry model to full-scale industrial operations.
Artificial intelligence has pushed scams to an unprecedented level, both in quality and quantity. Here is an overview of the most significant developments.
The Numbers Are Staggering
AI-enabled scams have increased by 1,210%, far outpacing the growth of traditional fraud (195%). Global losses tied to AI-assisted scams are estimated at $14.3 billion, and Deloitte projects that losses in the United States will rise from $12.3 billion in 2023 to $40 billion by 2027.
In France specifically, caller ID spoofing surged by 517% in 2025, and more than 500,000 victims have already received support after falling prey to AI-based scams such as voice cloning and fake financial advisers.
How AI Industrializes Scams
1. Highly Targeted Phishing
Generative AI is transforming phishing by creating highly personalized messages. Cybercriminals use personal data available on social media to craft emails tailored precisely to each victim.
2. Voice Cloning and Deepfakes
Voice-cloning technology now requires as little as three seconds of audio to reproduce a person’s voice with remarkable realism. By mimicking the voice or appearance of executives and public officials, criminals can deceive employees into authorizing fraudulent wire transfers—as demonstrated by the $26 million stolen in Hong Kong through a deepfake scheme.
3. Fraudulent Conversational Bots
AI-powered romance scams use large language models to sustain emotionally convincing conversations at scale. These bots can maintain dozens of simultaneous “relationships,” adapting their tone and personality to each target.
4. Fake Job Interviews
In January 2025, authorities dismantled a large-scale scam in which fraudsters used AI avatars to conduct fake job interviews, collecting sensitive personal information and confidential documents. Recruiting platforms report that identity-theft attempts involving AI-generated job postings have increased fivefold.
5. Large-Scale Cryptocurrency Scams
Check Point’s “Truman Show” operation uncovered a scheme involving 90 AI-generated “experts” deployed in controlled messaging groups to persuade victims to invest in fraudulent crypto platforms. According to Chainalysis, cryptocurrency scams caused $14 billion in losses in 2025, and AI-driven scams proved 4.5 times more profitable than conventional fraud.
Why Are These Scams So Hard to Detect?
AI-based scams remove many of the human limitations that once made social engineering easier to spot and slower to execute. There are no spelling mistakes, no foreign accents, and no obvious visual inconsistencies. People correctly identify AI-generated voices only about 60% of the time.
How to Protect Yourself
A few essential precautions:
- Verify any urgent request for money or sensitive information through a separate communication channel, even if the voice or face appears familiar.
- Be skeptical of offers that seem too good to be true, such as crypto investments or high-paying jobs with no genuine interview process.
- Establish a family “safe word” to confirm emergency calls from relatives.
- Never transfer funds solely on the basis of instructions received by email or phone call, even if they appear to come from a manager or supervisor.