What happened
Replaying the benefits-lure AiTM intrusion, August 2025
- 1
A benefits lure behind a bot check
From August 2025; still active in December 2025Benefits-themed emails from compromised marketing mailboxes and a bulk email service lead past a bot-check page to a look-alike Microsoft sign-in.
- 2
The Microsoft 365 sign-in relayed
From August 2025A reverse proxy relays the real Microsoft sign-in and captures the credentials and the session cookie, past MFA methods that are not phishing-resistant.
- 3
Sent on to a proxy of the Okta tenant
From August 2025For a user whose domain federates to Okta, the federation redirect is changed to a proxy of the user's own Okta tenant, which captures the single sign-on credentials and the Okta session cookie.
- 4
Okta sign-ins after the theft
Date not statedOkta observes sign-ins from cloud and proxy infrastructure after the theft, and an attempt to enroll an attacker-controlled device as an authentication factor.
A reverse-proxy phishing kit captured Microsoft 365 and Okta single sign-on credentials and session cookies; Okta observed successful sign-ins after the theft and an attempt to enroll an attacker-controlled device. Campaign emails reached hundreds of users at dozens of organizations; no victim organization is named and no count of accounts taken over is given.
Where provider and platform controls fit
In Microsoft Entra ID, a Conditional Access policy that requires the Phishing-resistant MFA authentication strength would have refused the relayed Microsoft 365 sign-in, though your user's password still reaches the proxy. The federation redirect is changed in your user's browser before any Microsoft session exists, so no Microsoft setting stops it. In Okta, an app sign-in policy that requires a phishing-resistant factor keeps a relayed sign-in out of the apps under it, and network zones bound fresh sign-ins from cloud and proxy networks, but not a replayed session cookie. The lure email stays outside the provider's reach.
Could this have been prevented?
In hindsight, not as a whole. 1 control would have refused a step outright, and 2 controls would have limited one. 2 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
- 2
- Contained
- 2
- 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 benefits-lure AiTM intrusion, August 2025
1 · A benefits lure behind a bot check
1 Not covered2 · The Microsoft 365 sign-in relayed
1 Prevented3 on other clouds3 · Sent on to a proxy of the Okta tenant
1 Contained1 Not covered4 · Okta sign-ins after the theft
1 ContainedKey facts
- Okta Threat Intelligence first published on October 21, 2025 and tracks the main activity as O-UNC-037, active since at least October 2025; Datadog Security Labs published on December 10, 2025 and said the campaign was still active
- Datadog traced a related variant to August 2025, and earlier campaign emails used other lure themes in October and November 2025
- The emails used compensation and benefits lures; some came from compromised third-party marketing mailboxes and others through a bulk email service, and an automated bot-check page came before the look-alike sign-in
- The kit relayed the real Microsoft sign-in in real time, captured the credentials and the session cookie, and could bypass MFA methods that are not phishing-resistant
- For users whose domain federates to Okta, the kit changed the federation redirect so the user reached a second-stage proxy of the user's own Okta tenant, which captured the single sign-on credentials and the Okta session cookie
- Datadog witnessed campaign emails sent to hundreds of users at dozens of organizations, named none and did not confirm any account takeover; Okta observed sign-ins from cloud and proxy infrastructure and an attempt to enroll an attacker-controlled device, with no count
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.Investigating an adversary-in-the-middle phishing campaign targeting Microsoft 365 and Okta usersDatadog Security Labs · Dec 10, 2025 · ResearchKit mechanics, sending infrastructure, gating, Microsoft 365 proxy, federation redirect to Okta, Okta session capture, scope, dates
- 2.Phishing campaigns use an employee-benefits lure to intercept Microsoft and Okta loginsOkta Threat Intelligence · Oct 21, 2025 · Vendor reportLure, reverse-proxy kit, federation redirect, credential and session capture, post-compromise authentication and device enrollment, defensive guidance
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.