Home / Technology / AI Chat Hijacked for Mac Malware Scams
AI Chat Hijacked for Mac Malware Scams
12 May
Summary
- Crooks exploit Claude's 'Shared Chats' feature for malware delivery.
- Fake 'Apple Support' ads on Google target Mac users.
- Malware avoids targeting Russian-language systems.

Cybercriminals have developed a sophisticated scheme to infect Mac users with malware, exploiting legitimate services like Claude and Google Ads. The campaign leverages Claude's "Shared Chats" feature, which allows users to create public links to AI conversations. Hackers craft fake conversations, impersonating "Apple Support," to promote fraudulent installation instructions for Claude Code.
These deceptive chats, often containing a warning about unverified content, are then advertised on Google. Users searching for Claude-related terms are directed to these malicious links, which display an authentic Claude URL. The embedded commands within the fake instructions trigger an infostealer infection on the user's device.
This operation notably avoids targeting systems with Russian language settings, indicating a deliberate effort to evade detection or a specific demographic focus. Security researchers have highlighted this tactic as a concerning new method for distributing malware.