Claude AI thwarts cyberattacks

Anthropic has successfully detected and neutralised attempts by malicious actors to exploit its Claude AI system for cybercrime. The AI model was being misused in several ways, including crafting phishing emails and generating malicious code. Hackers also attempted to use Claude to bypass existing safety filters.

These malicious activities included an influence-as-a-service operation using over 100 social media bots to manipulate political narratives. Another case involved scraping leaked credentials for IoT security cameras. Claude was also used in recruitment fraud schemes targeting Eastern European job seekers. Furthermore, the AI model was leveraged by novice actors to develop sophisticated malware. Anthropic has banned the accounts involved and is continuously upgrading its safeguards to prevent future misuse.