Description

Two advanced phishing attacks have been seen impersonating Microsoft 365 users using OAuth redirection weaknesses and brand spoofing techniques. These attacks evaded conventional security controls and were aimed at account takeovers (ATO). Attackers use established brands such as Adobe and DocuSign to deceive users into authorizing permissions for malicious OAuth applications. These applications, masquerading as "Adobe Drive," "Adobe Acrobat," and "DocuSign," lead victims to phishing sites that are intended for credential theft and malware distribution. In contrast to traditional phishing based on lookalike domains or email spoofing, this attack works solely within the Microsoft ecosystem, which makes it more difficult to detect. The attacks take advantage of the OAuth 2.0 authorization flow by controlling parameters such as 'response_type' and 'scope,' causing users to be redirected to attacker-controlled websites. This makes it possible for the phishing emails to evade security mechanisms like domain reputation screening, DMARC, and anti-spoofing. Because the phishing emails travel via Microsoft's actual servers, they will be less likely to set off alarms from conventional email security software. The hostile applications ask for minimal privileges, like "profile," "email," and "openid," which make them look legitimate but still allow attackers to achieve persistent access to delicate data. To reduce these risks, organizations need to implement phishing-resistant authentication technologies such as FIDO2 security keys and enforce strict conditional access policies. Legacy authentication protocols can be disabled, and number matching for MFA implemented to prevent attackers from circumventing security controls. Security teams also need to monitor Azure AD sign-in logs for anomalies, require phishing-resistant MFA, and provide continuous security training on OAuth consent phishing attacks to secure high-value employees and sensitive data.