Bithumb is running a security campaign to guide users on rules to prevent new financial scams that abuse artificial intelligence technology.

Bithumb said on May 14 it released a "Voice Phishing Complete Guide" episode as part of its Information Protection Day campaign. The content includes rules to prevent losses from deepvoice and deepfake-based voice phishing.

Voice phishing tactics have recently become more advanced, using AI to mimic in real time the voice and face of family members or officials at investigative agencies. In Hong Kong, a case occurred in which a victim was tricked by an AI-manipulated video conference into wiring about 40 billion won.

Impersonation attempts targeting digital asset investors are also increasing. The tactics include posing as exchange employees, financial institutions or family members to demand installation of remote-control applications, induce sharing of one-time passwords, or prompt transfers to a specific wallet address.

Bithumb stressed "three key security rules" to protect customer assets. It advised users not to share OTP codes and passwords under any circumstances and not to click links from unclear sources. It also said users should not comply with requests to send assets to a wallet address provided by someone else.

It also recommended actively using security features offered by exchanges, including two-channel authentication and blocking access from overseas IP addresses. It introduced everyday preventive measures as well, such as telecom carriers' AI phishing detection services and smartphone security settings.

A Bithumb official said crimes abusing rapid advances in AI are becoming more sophisticated, increasing investors' anxiety. The official said the company will continue security campaigns to create an environment in which users can trade with confidence.