Skip to content
All posts

M365 Device Code Phishing: Fake SharePoint Payment Lures

November 20, 2026 · Team PlayCISO
TL;DR · the short version

On 19 November 2026 Hornetsecurity's Threat Intelligence Lab reported that device-code phishing against Microsoft 365 had picked up again in November, using payment themes: payment advice, scheduled electronic payments, remittance and EFT notices, delivered as ZIP archives holding HTML or SVG files. The files redirected to counterfeit SharePoint-style workspaces that showed "payment documents" plus a code and sign-in instructions. A victim who typed the code into Microsoft's genuine device sign-in page and completed sign-in authorised a session the attacker had started, and Microsoft issued tokens to the attacker's client. No password is stolen and MFA does not stop it. Hornetsecurity names no actor or kit. Defenders should block the device code flow with Conditional Access wherever it is not needed, hunt the Entra sign-in logs for unexpected device-code sign-ins, inspect archive and HTML/SVG attachments rather than their names, and tell staff never to enter a code for a sign-in they did not start.

Put it into practice
Plan the move to phishing-resistant MFA
Device-code phishing gets past ordinary MFA. Plan which users and apps move to phishing-resistant methods first, and where device code flow can be blocked.
Open the phishing-resistant MFA planner

Payment-themed phishing is being used to make Microsoft 365 users approve sign-ins that attackers started. In a report published on 19 November 2026, Hornetsecurity's Threat Intelligence Lab describes two detailed campaigns and three more variants that send ZIP, HTML or SVG attachments dressed as payment advice, remittance or EFT notices. The files lead to counterfeit SharePoint-style pages that display a code and sign-in instructions. When the victim enters that code on Microsoft's genuine device sign-in page and completes sign-in, Microsoft issues tokens to the attacker's client. No password is stolen, and MFA completes as normal. Hornetsecurity names no actor or kit.

Who this is for: CISOs, identity and Microsoft 365 administrators, email security teams and finance teams who handle payment notices.

Based on Hornetsecurity's report and Microsoft's documentation, read on 20 November 2026. PlayCISO is not affiliated with Hornetsecurity or Microsoft. We publish no lure content or indicators; Hornetsecurity's report lists them.

Confirmed vs not yet known

Reported by Hornetsecurity: payment-themed device-code phishing campaigns in November 2026 using ZIP archives with HTML or SVG files, obfuscated redirectors, counterfeit SharePoint-style workspaces and Microsoft's real device sign-in page.

Not published: who is behind the campaigns, which phishing kit (if any) they use, how many organisations were targeted or how many accounts were taken over. Hornetsecurity gives none of these, and we do not add them.

How device code phishing works

The OAuth device code flow exists for devices that cannot show a browser, such as smart TVs, conference-room systems and command-line tools: the device shows a short code, and the user types it into Microsoft's sign-in page on another device. Device code phishing reverses who holds the device.

  1. The attacker starts a device-code sign-in from their own client and gets a code.
  2. The lure presents the code with a business pretext. Hornetsecurity puts it this way: "The attacker-controlled page supplies the code and the business pretext."
  3. The victim goes to the genuine Microsoft device sign-in page, enters the code and signs in, including MFA.
  4. Microsoft issues access and refresh tokens to the attacker's client, which can then reach the victim's mail, files and other services within the permissions granted.

Because the sign-in page is real, URL checks and "look for the padlock" advice do not help. Codes expire after a short time, which is why the lures generate them on demand; Hornetsecurity describes status and heartbeat endpoints that keep the lure page in step with its backend.

What Hornetsecurity found

  • Payment pretexts: payment advice, a scheduled electronic payment, a joint-check or updated statement, and EFT remittance notices.
  • Delivery: ZIP attachments holding an HTML or SVG file that acts as a click-triggered redirector, an HTML "securely shared payment advice" document, and one variant redirecting through Cloudflare Workers.
  • The landing page: a counterfeit SharePoint-style workspace showing remittance documents, a verification code and sign-in instructions.
  • Evasion: layered obfuscation (reversed fragments, substituted Base64 alphabets, repeating-key XOR and document.write loaders), which is why scanning attachments by name or extension misses them.

The technique is not new. Microsoft has written about earlier device-code operations, Storm-2372 (February 2025) and EvilTokens (September 2026), and on 17 November warned about the same risk for AI coding agents. Hornetsecurity does not link its November campaigns to either name, and neither do we.

Useful next, if this was relevant to you

Showing 1 of 3: CISO Readiness Scorecard

What security leaders should do this week

  • Block device code flow where nobody needs it. In Microsoft Entra Conditional Access, target the Authentication flows condition, select Device code flow and block access (Microsoft's policy guide). Audit legitimate use first (conference-room devices, some command-line tools), run the policy in report-only mode, then enforce with narrow exceptions.
  • Hunt for device-code sign-ins. In the Entra sign-in logs, filter Authentication Protocol for Device code and review successful sign-ins with no business reason: the application, account, source and time. Hornetsecurity notes the original sign-in method identifies the session even when later token renewals use another flow.
  • If someone was caught, treat it as an identity compromise. Revoke the user's sessions and refresh tokens, review mailbox rules, forwarding, OAuth app grants and recent file access, and check for payment-change requests sent from the mailbox. Work it with the identity compromise response checklist and, if invoices or payments were touched, the BEC response checklist.
  • Inspect what is inside attachments. Hornetsecurity's advice is to inspect the contents of archives and HTML attachments "rather than relying on their filenames", and to investigate payment-themed ZIPs containing interactive SVG or encoded document-sharing pages. Consider quarantining HTML and SVG attachments from external senders by default.
  • Give staff one simple rule. Only enter a device code for a sign-in or device setup you started yourself and understand. A document or payment message that hands you a code to "verify" is phishing; check through a trusted channel. Tell finance and accounts-payable teams first, since they are the target audience for these lures.
  • Move high-risk users to phishing-resistant sign-in and narrow token lifetimes for them. Plan it with the phishing-resistant MFA planner, and walk your team through the flow in the OAuth flow lab.

Frequently asked questions

What is device code phishing?

An attacker starts a legitimate Microsoft "device code" sign-in, the flow meant for TVs, printers and command-line tools, and gets a short code. The lure persuades the victim to type that code into Microsoft's real device sign-in page and finish signing in. Microsoft then issues tokens to the attacker's client. The victim never enters a password on a fake page, and their MFA completes normally.

Does MFA stop device code phishing?

Not by itself. The victim completes MFA on the genuine Microsoft page for a sign-in the attacker started, so the attacker receives valid tokens. Blocking the device code flow with Conditional Access where it is not needed, and teaching people never to enter a code they did not request, are the controls that address it.

What did Hornetsecurity find in November 2026?

Hornetsecurity's Threat Intelligence Lab reported on 19 November 2026 two detailed campaigns and three more variants that used payment themes (payment advice, scheduled electronic payments, remittance and EFT notices) delivered as ZIP, HTML or SVG attachments. The files redirected to counterfeit SharePoint-style workspaces that showed a code and sign-in instructions, so the victim would authorise an attacker-initiated device-code sign-in. It names no actor or kit.

How do I block device code flow in Microsoft 365?

In Microsoft Entra Conditional Access, create a policy that targets the Authentication flows condition, selects Device code flow and blocks access. Microsoft and Hornetsecurity advise auditing legitimate use first, testing in report-only mode and keeping narrowly scoped exceptions for the devices and accounts that genuinely need it.

How can I tell if someone in my tenant fell for it?

In the Entra sign-in logs, filter Authentication Protocol for Device code and investigate successful sign-ins with no expected business purpose, looking at the application, account, source and time. Hornetsecurity notes the original sign-in method identifies the session even when later token renewals use another flow. If you find one, revoke the user's sessions and refresh tokens and review mailbox rules, OAuth grants and file access.

Sources

Plan the move to phishing-resistant MFA
Device-code phishing gets past ordinary MFA. Plan which users and apps move to phishing-resistant methods first, and where device code flow can be blocked.
Open the phishing-resistant MFA planner
Take this with you
How device code phishing works · What Hornetsecurity found · What security leaders should do this week · Sources

Ready to practise the decisions these articles describe?

Run your first Cyber War Room free →

Pro