ConsentFix v3 Automates Azure Account Hijacking via OAuth Code Theft
Vulnerability Disclosure · 2 sources · Updated 5mo ago · First seen May 2, 2026

Researchers reported a new phishing technique, ConsentFix v3, that targets Microsoft Azure accounts by abusing the OAuth2 authorization code flow and trusted first-party Microsoft applications. The attack sends victims through a legitimate Microsoft sign-in process, then tricks them into pasting or dragging a localhost redirect URL containing an authorization code back into a phishing page. Attackers can then exchange that code for tokens, enabling access to Microsoft resources such as email and files without needing the victim’s password and even when MFA is enabled.
The latest variant expands on earlier ConsentFix methods by adding automation and scale through third-party services including Cloudflare Pages for phishing infrastructure and Pipedream for capturing authorization codes, exchanging them for refresh tokens, and collecting stolen tokens; reports also noted use of platforms such as DocSend. The technique also incorporates tenant enumeration and employee impersonation to improve targeting. Defenders were advised that mitigation is challenging because trust in Microsoft first-party apps is built into the architecture, but measures such as token binding, behavioral detections, and restricting app authentication can reduce exposure. Researchers said ConsentFix-style attacks have appeared in real campaigns, although widespread adoption of the v3 variant has not yet been confirmed.


