Advertisement

New botnet hijacks AI-powered safety software on Asus routers



Thank you for reading this post, don't forget to subscribe!

GreyNoise stated its in-house AI software, SIFT, flagged suspicious visitors aimed toward disabling and exploiting a TrendMicro-powered safety function, AiProtection, enabled by default on Asus routers.

Trojanizing the security web

Asus’ AiProtection, developed with TrendMicro, is a built-in, enterprise-grade safety suite for its routers, providing real-time menace detection, malware blocking, and intrusion prevention utilizing cloud-based intelligence.

After gaining administrative entry on the routers, both by brute-forcing or exploiting recognized authentication bypass vulnerabilities of “login.cgi” — a web-based admin interface, the attackers exploit an authenticated command injection flaw (CVE-2023-39780) to create an empty file at /tmp/BWSQL_LOG.

Doing this prompts the BWDPI (Bidirectional Net Knowledge Packet Inspection) logging function, a element of Asus’ AiProtection suite aimed toward inspecting incoming and outgoing visitors. With logging turned on, attackers can feed crafted (malicious) payloads into the router’s visitors, as BWDPI isn’t meant to deal with arbitrary knowledge.

On this specific case, the attackers use this to allow SSH on a non-standard port and add their very own keys, making a stealthy backdoor. “As a result of this secret’s added utilizing the official Asus options, this config change is persevered throughout firmware upgrades,” GreyNoise researchers stated. “In case you’ve been exploited beforehand, upgrading your firmware will NOT take away the SSH backdoor.”

Whereas GreyNoise didn’t specify a selected CVE used as an authentication bypass for preliminary entry, Asus just lately acknowledged a vital authentication bypass vulnerability, tracked as CVE-2025-2492, affecting routers with the AiCloud function enabled.