Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
OAuth consent phishing tricks a user into granting a malicious application permission to their mailbox and files, producing access that survives password resets and multi-factor authentication.

What Is OAuth Consent Phishing?
OAuth consent phishing, also called an illicit consent grant attack, persuades a user to approve a malicious application’s request for access to their data. No password is stolen, no multi-factor prompt is bypassed, and nothing is typed into a fake login page. The user authenticates on the genuine Microsoft or Google sign-in screen and then clicks Accept on a permissions dialogue. From that moment the attacker holds a refresh token that reads their mail and files independently of the account password.
OAuth 2.0 exists so that a third-party application can act on a user’s behalf without holding their password. The flow is familiar: a calendar tool asks to read your events, you approve, it receives a token scoped to that permission.
An attacker registers an application in their own tenant, configures it as multi-tenant so it can be consented to from anywhere, and requests permissions such as Mail.Read, Mail.Send, Files.ReadWrite.All, and crucially offline_access, which supplies a long-lived refresh token. They then send a link to the legitimate authorisation endpoint with their client ID in the query string.
Everything the user sees is real. The domain is login.microsoftonline.com, the TLS certificate is Microsoft’s, the sign-in uses their existing session, and the consent dialogue is the standard one. The only untrustworthy element is the application name and publisher, and attackers choose names such as “Document Sync”, “Office365 Secure Mail”, or the name of a real product the organisation uses.
The first mass example was the Google Docs worm of 3 May 2017, in which an application literally named “Google Docs” requested contact and mail permissions, then used each victim’s address book to spread. It reached around a million accounts in roughly an hour before Google revoked it.
Microsoft has documented repeated abuse since. In December 2022 it disrupted a campaign in which attackers impersonated verified publishers to make malicious applications look legitimate, a signal many organisations had been told to trust. In January 2024, Midnight Blizzard—the group also known as Nobelium—was reported to have gained access to Microsoft corporate mailboxes by password spraying a legacy test tenant and then abusing an OAuth application to extend that access. The technique appears throughout business email compromise investigations, where the grant provides durable mailbox reading for fraud reconnaissance.
The decisive control is removing the user’s ability to make the decision.
Revoke the service principal and all its tokens, not just the user’s sessions. Reset the password as well, but do not treat that as the remediation. Then establish what the application read during the window: mailbox audit logs, file access logs, and any mail forwarding or inbox rules created afterwards. Our Investigation and Incident Response Services cover cloud identity compromises where the evidence lives entirely in tenant audit logs.
No. The user completes a genuine sign-in at the real identity provider, including any multi-factor challenge, and the attack happens afterwards at the consent screen. Strong authentication confirms the right person is present; it says nothing about whether the application they are approving is trustworthy. Consent restrictions and an admin approval workflow are the controls that address this.
Because the attacker holds a refresh token issued to their application, which is independent of the user’s credential. Changing the password invalidates sessions but leaves the OAuth grant in place, so the application continues to obtain fresh access tokens. Remediation requires revoking the service principal or the specific permission grant, which is a separate administrative action.
Enumerate the enterprise applications and service principals in your tenant along with their granted permissions, then review anything unfamiliar, anything with high-impact scopes such as mail or file access, and anything consented to by a single user rather than an administrator. Pay particular attention to grants including offline_access, and check the creation date against known incident windows.
It works well when paired with an admin consent workflow, which lets users request an application and gives administrators a queue to review. Blocking consent without that workflow tends to push people towards personal accounts and unmanaged tools. A middle option allows consent to applications from verified publishers for a defined set of low-impact permissions only.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.