What happened
Replaying the verified publisher December 2022 intrusion
- 1
A verified badge on the actors' apps
Before Dec 6, 2022; date not statedFraudulent partner accounts that impersonate real companies add a verified publisher to app registrations the actors created.
- 2
Consent to read mail
Dec 6–27, 2022Lure files lead users to the Microsoft consent prompt, and users grant the apps delegated access to mail with offline access.
- 3
Mail read through the grants
Dec 6–27, 2022; individual dates not statedWith the delegated grants and long-lived refresh tokens, the apps are used to exfiltrate email without further user action.
Users in an unstated number of organizations, mainly in the UK and Ireland, granted three malicious apps delegated access to their mail, and Microsoft says the apps were used to exfiltrate email. Which mail was read is not stated.
Where a tenant setting fits
The verified badge came from Microsoft's partner program, which no setting of your tenant reaches, and the apps used the grants they were given. The decision that is yours is the consent itself. Microsoft Entra user consent turned off, with the admin consent workflow on, sends a user's request to an administrator instead of granting it. Allowing user consent for verified publishers does not refuse apps like these by itself, because their publishers were verified.
Could this have been prevented?
In hindsight, not as a whole. 1 control would have refused a step outright. 3 steps have no cloud-provider control. The counts below are control assessments, not attack stages: several can address the same step. The same replay with the controls in place before the incident shows where each phase would have been refused or bounded.
- 1
- Prevented
- 0
- Contained
- 3
- Not covered
How it could have been prevented
The same replay with the controls in place. Green marks where a step is refused, blue where it is bounded, and red where no control reaches it. Open a row to read the control, why it acts, and its caveat.
Replaying the verified publisher December 2022 intrusion
1 · A verified badge on the actors' apps
1 Not covered2 · Consent to read mail
1 Prevented1 Not covered3 · Mail read through the grants
1 Not coveredKey facts
- Proofpoint first observed the malicious apps on December 6, 2022, and the campaign ended on December 27, 2022, per Proofpoint.
- Three apps from three different malicious publishers, linked to the same infrastructure, targeted the same organizations, per Proofpoint.
- In two cases the publisher verification was granted one day after the app was created, per Proofpoint.
- The granted refresh tokens had an expiry of over a year in most cases, per Proofpoint.
- Microsoft became aware of the campaign on December 15, 2022 and says the actors used the apps to exfiltrate email.
- Microsoft disabled the apps across all tenants and the actors' partner accounts and notified the customers whose users had consented; the time of the disabling is not stated.
Sources
Every fact in this breakdown comes from the public reporting below. The replay is a model of that reporting, not a reconstruction of the environment it describes.
- 1.Dangerous Consequences of Threat Actors Abusing Microsoft's "Verified Publisher" StatusProofpoint · Jan 31, 2023 · ResearchPublisher impersonation, verification timing, the three apps, delegated permissions, lure files, user consent, token lifetime, campaign dates, Microsoft's disabling of the apps
- 2.Threat actor consent phishing campaign abusing the verified publisher processMicrosoft Security Response Center · Jan 31, 2023 · AdvisoryFraudulent Cloud Partner Program enrollment, verified publisher added to app registrations, email exfiltration, UK and Ireland targeting, disabling of apps and accounts, customer notification, recommended consent settings
- 3.Microsoft disables verified partner accounts used for OAuth phishingBleepingComputer · Jan 31, 2023 · NewsMicrosoft statement, customer notification
The estate in the replay is illustrative and drawn at the level of attack stages; the victim's real environment is not public. Verdicts describe what a control does when it is in place before the incident. They are not an assessment of the victim's security program.