Qianxin: Over 2,000 counterfeit DeepSeek websites, with 60% of IP addresses located in the United States

With the popularity of DeepSeek, the large-scale model sought after by the domestic AI company Depth, various counterfeit websites have also emerged, bringing security risks.

On February 6, it was learned from the Qianxin XLab that there are more than 2000 counterfeit DeepSeek websites and phishing websites, and the number is increasing rapidly.

Distribution of counterfeit domain name registration time

Analysis shows that 60% of the IP addresses resolving these counterfeit domains of DeepSeek are located in the United States, with the remaining primarily distributed in Singapore, Germany, Lithuania, Russia, and China, indicating a globalized feature of counterfeit domains.

Counterfeit domain name resolution results

According to the statistical analysis of domain name registrations from December 1, 2024 to February 3, 2025, a total of 2,650 websites imitating DeepSeek appeared during this period. Large-scale registration of counterfeit domain names began on January 26, 2025 and reached its peak on January 28. Although the growth rate has decreased since then, the number of counterfeit domain names continues to increase.

The main purposes of these counterfeit domain names are phishing fraud, domain name grabbing, and traffic redirection. Phishing fraud mainly involves stealing user login credentials, misleading users through similar domain names and interfaces, and deceiving users into purchasing virtual assets.

Qihoo 360 told the reporter of The Paper that some of these imitations may be for commercial purposes, intending to sell promising domain names or attract users by taking advantage of the popularity of DeepSeek. However, there are also many malicious actors who use similar domain names and interfaces to mislead users, and even spread malicious software, steal personal information, or deceive subscription fees.

These counterfeit websites use similar domain names and interfaces to mislead users, spreading malicious software, stealing personal information, or deceiving subscription fees. In addition, scammers closely follow technological trends, leveraging market excitement to introduce various high-end features of air coins (virtual currencies without substantial value) under the so-called “DeepSeek blessing,” and even claim to have websites where you can purchase DeepSeek’s original stocks.

This model is highly similar to the trend of a large number of counterfeits and frauds quickly appearing after many technological hot products (such as ChatGPT) became popular in the past, which may also cause users to suffer large financial losses.

When accessing DeepSeek related services, it is recommended that users must confirm that they are visiting the official website. For other domain names, unless the authenticity of their identity can be confirmed, it is not recommended to engage in Depth interaction, especially when it involves sensitive data such as usernames and passwords, which requires extreme caution.

On January 20th, DeepSeek, a Chinese AI startup, launched the large model DeepSeek-R1. As an open source model, R1’s performance in tasks such as mathematics, code, and natural language reasoning can rival that of the formal version of OpenAI o1 model, and it adopts the MIT license agreement, supporting free commercial use, any modification and derivative development, etc.

Due to its open source and low cost advantages, DeepSeek has become one of the hottest topics during this year’s Spring Festival. Data shows that within 20 days of its launch, the daily active users of DeepSeek exceeded 20 million. Several domestic cloud giants including Alibaba, Tencent, Huawei, and Baidu have announced their access to DeepSeek, forming a powerful ‘circle of friends’.

(Source: The Paper)

Source: Oriental Fortune Network

Author: The Paper