Product
Since September 2026, threat actors have leveraged fake Anthropic AI preview invitations, abused legitimate services, and utilized multi-layer CAPTCHA bypasses to harvest Google and Microsoft credentials.