Hackers Abuse Google Ads and Claude.ai Shared Chats to Target Mac Users

What Happened?

Security researchers discovered threat actors creating malicious Google Ads that promote fake AI applications designed to look legitimate. When users click these ads, they’re directed to websites hosting counterfeit software that mimics popular AI tools.

The attackers take their deception further by creating Claude.ai shared chat conversations that appear to come from trusted sources. These shared chats contain links to download the malicious applications, giving victims a false sense of security since the links appear to originate from Anthropic’s official platform.

Once installed, the fake applications deliver malware payloads to infected Mac systems. The malware can steal sensitive data, establish persistent access to compromised machines, and potentially serve as a gateway for additional attacks.

This campaign represents a significant evolution in social engineering tactics. By combining malvertising with abuse of legitimate AI platform features, the attackers create multiple layers of apparent credibility that can fool even cautious users.

The Impact

This attack highlights how cybercriminals adapt to exploit emerging technologies and trusted platforms. The use of Claude.ai’s sharing feature shows attackers studying legitimate tools to find ways to abuse their functionality for malicious purposes.

Mac users face particular risk because many believe Apple’s operating system provides inherent protection against malware. While macOS does include security features, this campaign shows determined attackers can still find ways to compromise systems through social engineering and user deception.

The incident also raises questions about how AI platforms can prevent abuse of their sharing features without limiting legitimate use cases. Companies hosting these services must balance accessibility with security to prevent similar exploitation.