In 2025, there were over 248 Web3 phishing incidents, with AI becoming a new assistant for attacks.

[Chain News] I just saw a report on the statistics of Web3 security from last year, and it's quite shocking. A total of 248 phishing incidents were recorded throughout 2025, with the total amount of money lost exceeding $700 million—this is just the cases that were reported.

The most frightening thing is not the phishing itself, but that the attacks have become smarter now. Hackers have started using AI tools that can quickly generate phishing pages identical to real websites and can also spoof wallet pop-ups. Not only that, they use AI to generate scam messages in multiple languages and then combine on-chain transaction data with social media information to precisely target users for their attacks - it's like transforming traditional net fishing into sniping.

What does this mean? Traditional defense methods are starting to fail. Those red flags identified by experience are becoming less effective. For us ordinary users, we need to be more vigilant. Project teams also need to keep up with the pace, not just conduct security audits, but also enhance user education and risk warnings at the operational level.

Calculating it out, if this trend of escalating attacks continues, the data for 2026 might be even more frightening. The same old saying applies—your own awareness of security is the last bastion in Web3.