Knight Office Phishing Kit Hijacks Microsoft 365 Sessions After MFA

Knight helmet over digital lock, cyber threat.

A new phishing-as-a-service operation called Knight Office is targeting Microsoft 365 and Google Workspace users by stealing active login sessions instead of passwords. The campaign can capture a session after a victim completes a legitimate-looking MFA sign-in, allowing attackers to access accounts without triggering many traditional password alerts.

Key takeaways

  • Knight Office uses fake DocuSign-style messages and device-login prompts.
  • Stolen sessions can let attackers bypass another MFA prompt.
  • Attackers may register rogue devices and passwordless credentials for persistence.
  • Small businesses should monitor post-MFA activity, not only failed logins.

How the Knight Office attack works

The campaign reportedly begins with an urgent email that appears to request a document signature. In some cases, the message is self-spoofed so it looks as though it came from the recipient’s own address. Other observed lures imitate voicemail notifications or shared-document alerts.

Links send victims through multiple redirects, including legitimate services and compromised websites, before reaching a counterfeit Microsoft sign-in page. The page presents a device-login code that resembles Microsoft’s genuine sign-in process. When a user enters the code, signs in, and approves MFA, the attacker’s infrastructure captures the resulting active session.

Because the attacker uses the authenticated session, rather than guessing a password, the activity may not generate conventional failed-login or password-spray alerts. For a small business without dedicated security staff, this can make routine cloud-account monitoring especially important.

Attackers can establish lasting access

Huntress investigators found evidence that the operator registered an attacker-controlled device in the victim’s Microsoft Entra ID environment, formerly known as Azure Active Directory. The attacker also attached a Windows Hello for Business credential to the compromised account.

That step can provide a route back into the account after the original session is revoked. It also demonstrates why account recovery should include a review of newly registered devices, authentication methods, and passwordless credentials—not just a password reset.

Helpdesk.Live customers can use managed endpoint security, cloud administration support, and human IT help to make these reviews part of a practical response process. AI support can help explain alerts and next steps, while live-agent assistance is available through the applicable service option when hands-on guidance is needed.

Campaign appears broader than isolated incidents

The same attacker console was associated with at least nine confirmed phishing attacks against Microsoft 365 and Google Workspace accounts in one customer group over two weeks. More than 700 similar messages had also been reported through a security-awareness platform since April, indicating that the operation may have been active for months.

Researchers identified an attacker dashboard with visitor statistics, a bot check, and live access to harvested information. They also observed lookalike domains and minor spelling changes, such as replacing the letter “i” with a lowercase “l,” likely to evade simple keyword-based filters.

What small businesses should do now

Business owners and administrators should treat successful MFA approval from an unexpected message as a potential security event. Recommended actions include:

  1. Review sign-ins occurring after MFA from unfamiliar devices, locations, or hosting providers.
  2. Inspect Entra ID for unrecognized devices and newly added Windows Hello credentials.
  3. Revoke suspicious sessions and authentication methods, then reset affected accounts.
  4. Check mailbox rules, forwarding settings, OAuth grants, and recent account activity.
  5. Train staff to avoid entering device codes from unsolicited emails or documents.

Proactive monitoring and managed threat detection can help surface unusual activity earlier, while clear escalation paths reduce confusion during an incident. Helpdesk.Live’s flexible model combines remote monitoring, security tools, AI assistance, and optional live-agent support so smaller organizations can select the level of help that fits their needs.

Sources

Scroll to Top