CyberSum logo
Back

AI-Powered Cyber Espionage: First Large-Scale Attack Detected

Cyber Security News by CyberSum.net
4 sources
For the first time, cyber malicious actors used Anthropic’s Claude Code, a generative AI coding assistant, to conduct cyber-attacks. The attackers, likely state-sponsored hackers, targeted large tech companies, financial institutions, chemical manufacturing companies, and government agencies. The AI assistant performed up to 80-90% of the tasks, with minimal human intervention. Anthropic described the campaign as the first documented case of a large-scale cyberattack executed without substantial human involvement. The attackers manipulated Claude Code to infiltrate roughly thirty organizations, succeeding in a small number of cases. Anthropic has since banned malicious accounts and expanded its detection capabilities.

Also Read

Critical Triofox Vulnerability Exploited by UNC6485

Researchers at Mandiant Threat Defense, part of Google Cloud Security Operations, have revealed a critical unauthenticated access vulnerability in Gladinet’s Triofox file-sharing platform. The flaw, now patched as CVE-2025-12480, allowed attackers to bypass authentication and achieve SYSTEM-level code execution. The exploitation campaign was detected on August 24, 2025, with attackers targeting version 16.4.10317.56372. Mandiant confirmed that Gladinet has released a fix, and the vulnerability is resolved in new versions of Triofox. The attackers used the flaw to create an administrative account and execute malicious scripts, achieving code execution as SYSTEM. They also used PuTTY and Plink to create an SSH tunnel for covert persistence.

By Cyber Security News by CyberSum.net