Skip to content

Response, New York. Noah Park: Check Point Patches Actively Exploited Management Server Path Traversal. Detection, Dallas. Ava Okello: Report Finds Nearly Half of Deployed Detections Need Attention Before SOCs Can Trust Them.

SOCtember

Always First. Fast SOC News.

Huntress graphic headed M365 Token Spray and Replay Campaign by the Numbers, showing 200,000+ login attempts, 79 organizations affected, 97 accounts compromised, 497 successful sign-ins and 48 rogue Entra devices registered.
Graphic: Huntress, M365 token spray and replay campaign by the numbers, Oct 22, 2026 (fair use).

Response

Huntress Says a Token Spray Campaign Replayed Stolen Microsoft 365 Sessions Into 79 Organizations, Registered 48 Rogue Entra Devices and Hid Payroll Mail Behind One-Letter Inbox Rules

Over 10 weeks, an operator captured post-MFA sign-in sessions through adversary-in-the-middle relays and device-code phishing, then replayed them from a rented block of German hosting addresses. Huntress counted 497 successful sign-ins against more than 200,000 failed attempts. A password reset alone does not remove the rogue devices and authenticator apps the attacker left behind.

Noah Park, ResponseNew York6 min read

NEW YORK - An operator who never needed a password signed in to Microsoft 365 accounts at 79 organizations over 10 weeks this summer by replaying sign-in sessions stolen after the victims had already passed multifactor authentication, according to research the security company Huntress published Thursday.

Huntress said it recorded 497 successful sign-in events between July 14 and Sept. 22, 2026, and traced more than 200,000 failed login attempts from the same window to a single block of addresses, 193.26.115.0/24, owned by 1337 Services GmbH, a Hamburg hosting provider also known as AS210558. Its own summary graphic counts 97 compromised accounts. Microsoft Entra ID Protection independently raised verified threat actor IP detections against four of the addresses in multiple customer tenants, Huntress said.

The company calls the technique token spraying: the same pattern as password spraying, but with stolen tokens instead of guessed passwords. The tokens came from adversary-in-the-middle relays sitting behind fake sign-in pages, from device-code phishing and from social engineering during a legitimate login. Because the token is minted after MFA, whoever holds it does not have to pass MFA again.

Capture on one network, replay from another

A July 21 case shows the speed. At 17:18:20 UTC a relay at 185.238.2.195 caught the victim at the MFA prompt. At 17:30:21 the session was hijacked from 193.26.115.199. At 17:43:15 the same session was in use from a second 1337 Services address, 124.198.132.33, and at 18:08:50 a new authenticator app labeled "iPhone 13" was activated on the account. On Sept. 22 the gap between the victim reaching the MFA prompt and the replay was four minutes.

Huntress chart titled IP Addresses Used in Token-Spraying Campaign, showing six addresses in 193.26.115.0/24 used one after another between July and September 2026.
Chart: Huntress, the six replay addresses in 193.26.115.0/24 and when each was active, Oct 22, 2026 (fair use).

The operator used six addresses in the /24, one after another, often after a quiet spell. Microsoft flagged them individually, not as a range. "Blocking any single address would have stopped one phase at most," Huntress wrote. The split also matters for containment: the relays that capture sessions and the hosts that replay them sit on different providers, so blocking one half does not stop the other.

Huntress saw two toolsets on the replay hosts. Build A, a scripted replay with the user agent node, traded tokens through the Microsoft Office and Microsoft Authentication Broker clients against Microsoft Graph, Exchange Online, SharePoint Online and the Device Registration Service. Build B replayed browser sessions with a current Chrome or Edge user agent through OfficeHome.

Huntress table comparing Build A scripted token replay and Build B browser session replay by hosts, user agent, request type, client app, target and device, including the rogue Entra-joined device named DESKTOP- plus eight hex characters.
Screenshot: Huntress, the two replay builds and their sign-in fingerprints, Oct 22, 2026 (fair use).

Persistence that outlives a password reset

Across Build A sign-ins, Huntress counted 48 distinct Entra-joined devices named DESKTOP- plus exactly eight uppercase hexadecimal characters, such as DESKTOP-54521F46. Windows' own out-of-box default is DESKTOP- plus seven alphanumeric characters. A registered device can give the attacker a more durable way to request tokens than the stolen session, and it stays in the directory until someone deletes it. In 19 organizations the victim's existing MFA methods were wiped and a new method registered within three hours of the first sign-in from a campaign address.

On Sept. 11 the operator switched to 193.26.115.72 and Huntress sent incident reports to 22 customers within two hours. That address logged 1,052 blocked sign-ins, all of them in four organizations where Conditional Access was doing its job. The organizations where it got in were different ones. "Do not trigger response based on spray volume, as it points at the most-sprayed, least-compromised tenants; prioritize by credential validity instead," Huntress advised.

Then came the mail. Every suspicious inbox rule Huntress found was created from a VPN or rented hosting address, never from 1337 Services, and the names were built to be forgettable: a single period, sss, iT, a row of dots. In one case the attacker modified a rule named s to catch any message mentioning payroll, direct deposit, bank or bonus, then five minutes later, from the Interserver address 167.148.41.96, created a rule named d that marked all mail from an outside CPA as read and moved it to Conversation History. The compromised mailbox was used to send that CPA a payroll-related email.

Huntress table of hidden inbox rules named period, sss, ss, iT, comma-period, a row of dots, s and d, with actions such as mark as read and delete or move to RSS Feeds, RSS Subscriptions or Conversation History, and triggers including payroll terms and one external CPA's address.
Screenshot: Huntress, inbox rules created on compromised mailboxes, Oct 22, 2026 (fair use).

Device-code phishing showed up in the August wave, a technique Microsoft dissected in its September report on the EvilTokens kit. Failed replays were informative too: attempts to reuse a session after a password change came back with Microsoft's SsoArtifactRevoked error, AADSTS50133.

What responders should do now

  • Revoke sessions and reset the password. Huntress notes that a reset alone does not invalidate an access token, which has a time to live of approximately 90 minutes, and does not remove attacker-registered devices or MFA methods.
  • Audit Entra devices named DESKTOP-, Device- or Work Pc plus eight hex characters, especially ones registered from hosting addresses. The pattern is rare but not unique to this campaign, so confirm before deleting.
  • Audit MFA methods for ones cleared and re-registered within minutes, and remove any authenticator the user does not recognize, including a device labeled iPhone 13.
  • Review inbox rules with one- to three-character or punctuation-only names, rules that move mail to RSS Feeds, RSS Subscriptions or Conversation History, and rules keyed on payroll or banking terms. In four mailboxes, Huntress found rules created through the Outlook REST API rather than Outlook itself.
  • "Treat Conditional Access blocks (AADSTS53003) as a sign of a valid credential, not a save," Huntress wrote. The block happens after authentication succeeds.
  • Hunt for OAuth2:Token and Cmsi:Cmsi sign-ins from hosting infrastructure that use the Microsoft Office or Authentication Broker client with a node or bare Mozilla/5.0 user agent, focusing on sign-ins that touch mailboxes or the Device Registration Service.
  • Restrict device-code flow with Conditional Access authentication-flow policies, require MFA or block for all users, all cloud apps and all client app types, and stop standard users from consenting to new enterprise applications.
  • Sweep sign-in logs for 193.26.115.0/24 plus 124.198.132.33, 45.83.31.44 and 192.159.99.172, the relay 185.238.2.195 and the inbox-rule address 167.148.41.96.

Sources:


Noah Park covers incident response, containment, and blue-team operations for SOCtember from New York.

Related stories

Response desk