Cybersecurity researchers have uncovered a months-long campaign that broke into Microsoft 365 accounts at 79 organizations without ever needing a victim's password. Instead of guessing credentials, the attackers stole the login tokens issued after multi-factor authentication and replayed them at scale — a technique the researchers call "token spraying."
Security firm Huntress published its findings on October 22, after investigating suspicious sign-in activity across its customer base. The campaign ran from July 14 to September 22, 2026, and produced 497 successful Microsoft 365 sign-in events across 79 victim organizations, alongside more than 200,000 failed login attempts traced to a single block of internet addresses belonging to German hosting provider 1337 Services GmbH.
Tokens minted after MFA, replayed at scale
The attack starts with deception, not code-breaking. Victims landed on fake sign-in pages backed by adversary-in-the-middle relays, which passed the genuine Microsoft login page and its multi-factor prompt through to the victim while quietly keeping the session token Microsoft issues once sign-in completes. Because that token is created after the multi-factor step, whoever holds it can use it without passing multi-factor authentication again.
Other tokens came from device-code phishing, in which victims were tricked into entering a code on Microsoft's legitimate sign-in page, and from social engineering during an ordinary login. The attackers then replayed the stolen tokens from their own machines in automated sign-in attempts against Microsoft 365 services. Huntress coined the term "token spraying" for the pattern: the same high-volume approach as password spraying, but with stolen tokens instead of guessed passwords.
Speed was part of the danger. In one incident on July 21, a hijacked session appeared from an attack address just 12 minutes after the victim reached the multi-factor prompt; on September 22, the gap was four minutes. Once inside, the intruders set the stage for business email compromise by registering rogue devices, adding new authenticator apps and creating hidden inbox rules.
Some context matters for the scale. The figures cover Huntress's own customer environments, not every Microsoft 365 tenant, and 497 sign-in events do not necessarily mean 497 separate compromised accounts. Microsoft's Entra ID Protection independently flagged several of the replay addresses as linked to nation-state actors or cybercrime groups, though neither company has named a specific actor. Huntress also notes that 1337 Services is a legitimate hosting company, and that identifying its addresses concerns the servers used in the attacks, not participation by the provider.
Four phases, one IP block
Over the course of the campaign, the attacker moved between six addresses in the 193.26.115.0/24 range across four phases, with breaks of a week or two between some of them. Huntress says blocking any single one of those addresses would have stopped a single phase at most, which makes simple IP blocklists a weak defense on their own.
What to do if you run Microsoft 365
The practical fix starts with authentication that cannot be relayed. Phishing-resistant methods such as FIDO2 security keys and passkeys are not vulnerable to the relay trick, because there is no token for an attacker to capture and replay. Administrators should also review sign-in logs for unfamiliar locations and devices, revoke suspicious sessions, and hunt for the campaign's telltale persistence signs: unexpected devices, new authenticator registrations and inbox rules nobody remembers creating.
Restricting or disabling the device-code flow where it is not needed, tightening Conditional Access policies around token protection, and acting on Entra ID Protection alerts complete the picture. The campaign is a reminder that multi-factor authentication raises the bar but does not end the game: once a session token leaks, the password and the second factor have already done their work, and the token itself becomes the key.



