Crude Anthropic Phishing Bypasses Top Email Gateways
A crude Anthropic-themed phishing campaign is successfully evading some of the industry’s most recognized email security products. Since September, the operation has distributed poorly-crafted emails claiming to offer early access to a new Anthropic AI social media tool, targeting Google and Microsoft account credentials. The emails contain obvious red flags: grammatical errors, sender typos, missing brand impersonation, and a confusing authentication flow that even breaks down mid-attack.
The campaign’s success despite its low quality stems from three tactical choices. The attacker leverages a legitimate GoTo Webinar sender address to bypass sender reputation checks, masks the malicious registration link through a URL shortener, and uses dual CAPTCHAs—one from Cloudflare to block automated scanning and another designed to detect VPNs and Tor browsers used by human defenders or security tools. These modest techniques were sufficient to penetrate defenses from Proofpoint Secure Email Gateway and Microsoft Defender for Office 365.
Security leaders increasingly acknowledge that email gateways alone cannot stop threats and recommend shifting strategy toward identity-centric defenses. Recommended controls include strict least-privilege policies, phishing-resistant multifactor authentication, privileged access management, and centralized password managers that validate domains before autofill. The campaign’s possible use of AI generation—evidenced by multiple phishing page variations and slightly malformed spoofing—suggests attackers can now deploy high-volume attacks with minimal manual effort.
Read the full article
Continue reading at Dark Reading →This is an AI-generated summary. Read the original for the full story.