How EvilTokens Turned Stolen Inboxes Into Fraud Playbooks
Microsoft explains how device code phishing authorizes an attacker's session.
Prepared automatically from the linked sources. The category byline does not confirm personal authorship or review. Editorial policy.
Microsoft’s September 22 account of its EvilTokens disruption describes a service that turned stolen email access into preparation for financial fraud. Its chatbot could examine correspondence, identify trusted relationships and suggest impersonation strategies. Canada was among the countries with the highest concentrations of victim activity observed by Microsoft. The company says it and partners seized websites and disabled supporting domains. Those actions disrupted infrastructure; they do not establish that every affected organization has recovered.
The service brought account compromise, target selection and fraud preparation together in a commercial interface, according to Microsoft’s account. Microsoft.
Editorial analysis: The revealing feature is what happened after access was obtained. An inbox supplied context about people, invoices and authority, beyond addresses alone. A fraudulent instruction could therefore borrow details from a real working relationship. That makes the origin of a message only part of the payment question. The reported capabilities explain why a familiar conversation can become useful to an attacker, but they do not tell us how any particular recipient responded. Microsoft.
What codes authorize
The technical report describes an entry route that required no password disclosure. Device code authentication normally lets someone sign in on equipment with a limited interface by entering a code through another device. In the phishing version, the attacker starts the request and supplies the code. The person entering it then authorizes the attacker’s session while using a legitimate authentication flow. Microsoft says this separation can circumvent traditional multifactor authentication protections. The report also describes deceptive lures based on invoices, requests for proposals and shared files. Microsoft’s technical report.
Stolen tokens supported mailbox access and persistence. Malicious inbox rules could conceal communications, and some tokens could grant additional devices access. Microsoft recommends blocking device code flow where possible, with narrowly scoped exceptions where needed. Editorial analysis: A legitimate sign-in page is therefore insufficient evidence that the request behind it belongs to you. The practical question is which session you are approving and whether you initiated it. For an organization, deciding where this authentication flow is necessary is a control question, alongside staff awareness. Microsoft’s technical report.
Where crypto enters
An April 2021 IC3 alert supplies historical context for cryptocurrency. It described business email compromise involving doctored wire instructions and funds directed toward cryptocurrency exchanges. In one route, a transfer reached an exchange’s account at a traditional financial institution. Alternatively, money passed through an intermediary first. Victims in both scenarios were unaware that funds were destined for cryptocurrency conversion. This is background on an established fraud pattern, not evidence of transactions carried out through EvilTokens. IC3’s 2021 alert.
Editorial analysis: The crypto connection need not be visible in the payment instruction a victim sees. Checking only whether a request mentions a wallet would miss the indirect route described in the alert. Its relevance is the mismatch between the apparent purpose of a payment and its eventual destination. For someone who discovers fraud, IC3 advised immediately contacting their financial institution to request a recall. That instruction is a response step, not a promise that transferred money can be recovered. IC3’s 2021 alert.