According to CoinMeta, Bitget Wallet announced the enhancement of AI risk monitoring. By analyzing web page code, domain name characteristics, and their similarity to legitimate websites through AI, the system identifies suspicious websites that are not listed in the known threat database. Based on its internal data, in October 2026, among the newly blocked risk domains, its self-developed AI contributed to 90% of the identifications, which is higher than the 36% in the second quarter. Nearly 70% of these domains were marked by AI without any corresponding alerts from external security intelligence databases, indicating that AI provides additional identification capabilities for the existing risk list. Bitget Wallet's internal monitoring in the third quarter showed that phishing-related signatures accounted for 57% of the four types of risk detections, followed by malicious addresses (17%), high-risk contracts (14%), and malicious authorizations (12%). As wallet usage expands into payments and daily finance, Bitget Wallet plans to further expand the risk monitoring capabilities driven by AI and explore the use of AI under human supervision to assist in security incident responses, in order to identify new types of risks and speed up response times.