A sophisticated malvertising campaign is currently abusing Google Ads to target users searching for Anthropic's Claude.ai. According to BleepingComputer, attackers are using legitimate-looking 'Claude mac download' ads that redirect to instructions for installing malware. This campaign is notable for its use of shared Claude.ai chat links to host malicious instructions, effectively leveraging the platform's own reputation to bypass browser security filters. This represents a 'trust-chain' attack where the victim's familiarity with AI tools is used to lower their defensive posture. The malware delivered is designed to exfiltrate browser cookies and keychain data, specifically targeting developers and AI researchers who are high-value targets for corporate espionage. This trend correlates with the 'AI Framework Subversion' campaign, as threat actors move from attacking the AI itself to attacking the users and infrastructure surrounding it.