Category: Finance | Title: Possessed Face: AI Face Swap Scams, Deepfake Fraud, and Regulatory Response in 2024 | Tag: AI Fraud | Meta Description: Facts on possessed face deepfake scams, financial losses, detection tools, and SEC enforcement actions in 2024...
What Is a Possessed Face in Digital Fraud
A possessed face refers to a real human face digitally superimposed onto another body or video using artificial intelligence, often for scams. In 2024, the FBI and FTC report a sharp rise in possessed face incidents used in business email compromise, romance fraud, and fake customer service calls. The term is linked to deepfake technology, which uses generative adversarial networks and diffusion models to create convincing synthetic media. The Global Anti-Scam Alliance estimated that online fraud losses exceeded $1.03 trillion globally in 2023, with AI-generated content playing an increasing role. For background on how these tools work, see the overview at https://www.forbes.com/sites/bernardmarr/2024/01/15/deepfakes-explained-what-they-are-and-how-they-work/.
Financial institutions and regulators now classify possessed face attacks as a subset of identity fraud. In 2024, banks reported that voice and video impersonation attempts grew by over 300 percent year on year. The U.S. Securities and Exchange Commission has flagged deepfake audio and video as a new vector for market manipulation and retail investor fraud. The European Union's AI Act, which entered force in August 2024, requires watermarking and labeling of synthetic media. The FTC also updated its Endorsement Guides in 2024 to address AI-generated personas used in ads.
How Possessed Face Scams Work in Finance
Common Attack Vectors
Scammers use open-source deepfake tools and commercial face-swap apps to create a possessed face that mimics a CEO, family member, or support agent. In a typical business scam, the attacker sends a video call or recorded message using a possessed face to authorize a wire transfer. According to a 2024 report by Sumsub, deepfake fraud attempts in the financial sector rose by 1,740 percent in 2023 compared with 2022. The average loss per incident reported by the FBI's Internet Crime Complaint Center in 2023 was over $10,000, though CEO fraud cases can exceed $100,000.
Detection and Prevention Tools
Companies such as Microsoft, Intel, and Sumsub have released real-time deepfake detection platforms that analyze micro-expressions, blinking patterns, and audio-video sync to flag a possessed face. Microsoft's Video Authenticator and Intel's FakeCatcher are used by some banks to screen customer-facing video calls. The SEC's Division of Enforcement has brought cases against firms that failed to implement controls against AI impersonation. For a technical overview of detection methods, see https://www.forbes.com/sites/bernardmarr/2024/01/15/deepfakes-explained-what-they-are-and-how-they-work/.
Regulatory and Market Response to Possessed Face Fraud
SEC and Global Enforcement
The SEC charged multiple entities in 2023 and 2024 with using AI-generated personas and deepfake audio to manipulate stock prices and deceive investors. In one case, the SEC obtained an emergency order to freeze assets tied to a scheme that used a synthetic executive voice to promote a penny stock. The U.S. Department of Justice has also pursued criminal charges under wire fraud and identity theft statutes for deepfake-enabled schemes. The UK's Financial Conduct Authority has issued guidance requiring firms to verify identities using liveness detection and multi-factor authentication.
Industry Standards and Technology Adoption
Major platforms including Tesla, SpaceX, and Meta have updated their internal security protocols to detect and block deepfake content in internal communications and public channels. Tesla and SpaceX have publicly warned employees