A new report reveals that 87 deepfake scam rings were dismantled across Asia in Q1 2025, highlighting growing concerns over AI-driven fraud operations.
The seriousness of deepfake scams is highlighted in a recent paper by Bitget, SlowMist, and Elliptic, which calls on people and organizations to take stronger precautions.
According to a recent analysis, deepfake technology is driving a boom in complex frauds, and the rise of artificial intelligence is contributing to a dramatic increase in crypto-related fraud.
According to the 2025 Anti-Scam Research Report, which was co-authored by blockchain security company SlowMist, analytics company Elliptic, and cryptocurrency exchange Bitget, at least 87 scam rings employing AI-generated deepfakes were taken down in the first quarter of 2025.
According to the research, cryptocurrency frauds totaled $4.6 billion in 2024, a 24% rise from the previous year. Deepfake technologies were used in nearly 40% of high-value fraud cases, and scammers are increasingly deploying complex impersonations of platform executives, founders, and famous personalities to trick users.
The CEO of Bitget, Gracy Chen, stated that “deepfakes have a unique advantage in both reach and believability due to the speed at which scammers can now generate synthetic videos and the viral nature of social media.”
Deepfakes present fresh dangers.
The research describes the structure of contemporary cryptocurrency scams, identifying three main types: social engineering schemes, Ponzi-style frauds masquerading as decentralized finance (DeFi) or GameFi initiatives, and AI-generated deepfake impersonations. It cautioned that deepfakes are particularly challenging to identify.
AI can simulate voice messages, text messages, activities, and facial expressions. For instance, using Telegram, X, and other social media platforms to falsify video endorsements of investment platforms from well-known individuals like Elon Musk and the prime minister of Singapore are strategies used to take advantage of public confidence.
Even real-time behaviors can be simulated by AI, which makes it harder to distinguish these frauds from the actual thing. In a May 13 post on X, Sandeep Narwal, co-founder of the blockchain platform Polygon, raised the alarm by disclosing that malicious actors had been using Zoom to pose as him.
On Telegram, he said, several people had contacted him, wondering if he was on a Zoom conversation with them and whether he was asking them to install a script.
SlowMist CEO Yu Xian advised users to confirm the authenticity of Zoom URLs and domain names to prevent becoming victims of video-based impersonation schemes.
Smarter defenses are needed in light of new scam threats.
Security experts say awareness and caution are essential as deepfake schemes become more realistic. According to the research, institutions must have robust technology defenses and regular security training. Companies are encouraged to protect email systems, undertake phishing simulations, and watch for code leaks. According to the report, establishing a security-first culture where employees verify before they trust is the simplest method to prevent fraud before it begins.
Chen provided a simple method for regular users: “Verify, isolate, and slow down.” She also said:
“Always verify information through official websites or trusted social media accounts—never rely on links shared in Telegram chats or Twitter comments.”
Chen also emphasized the importance of using several wallets for risky activities when experimenting with new platforms.
