AI-driven crypto scams in Q1 2025 resulted in over $200 million in losses, utilising deepfake technology and exploiting token approvals. Scammers deceive users through tasks promising rewards, using micro-rewards to build trust before draining funds. Experts warn users to limit approvals and employ safety measures. Deepfakes complicate fraud further, with prominent figures being impersonated to promote fake schemes.
In Q1 2025, losses due to AI-driven crypto scams exceeded $200 million, driven by sophisticated deepfake and blockchain tactics aimed at deceiving users. Reports by GoPlus Security highlight a concerning trend where attackers exploit both technical vulnerabilities in token approvals and psychological elements derived from deepfake media, ultimately resulting in substantial financial losses for crypto investors.
Detailed analyses reveal a multi-step scam tactic involving the deceptive release of seemingly legitimate cryptocurrency tokens that promise stable USDT rewards for completing simple online tasks. Scammers initiate trust by sending micro-rewards in small amounts over several days, which diminishes user suspicion. Once trust is established, victims are led to grant permissions for token approvals to externally owned accounts (EOAs), making them vulnerable to quick fund drainage by automated systems.
Experts, including ZachXBT, recommend that users should exercise caution by limiting token approvals and employing tools to revoke dormant permissions. It’s essential to check for the legitimacy of contracts, restrict the scope of approvals, and monitor transaction activities closely to mitigate the risk of theft. As crypto and DeFi tools become increasingly intricate, fraudsters blend technical deception with social engineering to create convincing fronts.
Additionally, the rise of deepfakes presents a new frontier in fraud, as scammers generate realistic video content featuring prominent figures endorsing fictitious projects. A notable incident features Binance co-founder Changpeng Zhao alerting his followers to false deepfake endorsements. In Ghana, a deepfake of Ashesi University President Patrick Awuah Jr. was used to promote a fake scheme, prompting the university to publicly distance itself from the fraudulent activities.
Recent findings from McAfee suggest that the average American encounters three deepfake videos daily, heightening the stakes concerning crypto fraud. The company has identified key characteristics of AI-driven scams, including urgency, celebrity endorsements, and unrealistic returns. The rapid advancement of generative AI in creating convincing content allows scammers to execute increasingly sophisticated and scaled campaigns, making it essential for individuals in the crypto space to remain vigilant against these evolving threats.
