MFA Was Enabled —
And They Still Got In.
A Real Microsoft 365 Attack Breakdown
A client's Microsoft 365 account was compromised despite having Multi-Factor Authentication enabled, a clean PC, and no visible malware. Here is exactly how it happened, why your phone may be the weakest link, and what every UAE business must do right now.
The Incident: What Actually Happened
The call came in like dozens of others: "Our Microsoft 365 account was hacked. But MFA was on. And Malwarebytes found nothing on the PC. How is this possible?"
The answer is not reassuring — but it is knowable. When we examined the Azure AD sign-in logs, what we found pointed clearly to one of the most sophisticated and increasingly common attack patterns in 2026 targeting Microsoft 365 security in UAE — businesses in Dubai, Abu Dhabi, and Sharjah alike: an Adversary-in-the-Middle (AiTM) phishing attack that stole an authenticated session token after MFA was already completed.
The attacker did not crack the password. They did not bypass MFA in the traditional sense. They let the legitimate user complete MFA — and then stole the resulting session token (session token theft), replaying it to access the mailbox from foreign IP addresses with no further authentication challenge required. This is why phishing-resistant MFA methods like FIDO2 and passkeys matter.
In this incident: MFA was enabled ✅ — No malware on PC ✅ — Successful logins recorded from multiple foreign countries ✅. This combination of facts is a near-textbook signature of an AiTM phishing attack or MFA fatigue approval. The device most likely compromised: the user's Android smartphone.
How Do Attackers Bypass MFA? The 6 Real Methods
MFA is not broken — but it is not invincible either. Here are the six primary methods attackers use to defeat MFA, ranked by likelihood in this specific incident:
1. Adversary-in-the-Middle (AiTM) Phishing
🔴 MOST LIKELY in this incidentThis is the number-one modern method for bypassing Microsoft 365 MFA. The attacker deploys a reverse proxy phishing page — a site that looks exactly like the Microsoft login portal but sits between the user and Microsoft's real servers. Here is the attack flow:
Phishing link delivered (WhatsApp, SMS, or email)
User receives a convincing link — often disguised as a Microsoft notification, shared document, or Teams invite.
User enters credentials on the fake proxy page
The page relays everything to Microsoft's real server in real time — including the MFA challenge.
User completes MFA — attacker captures session token
The authenticated session cookie is intercepted by the attacker's proxy before being passed to the user's browser.
Attacker replays token — full account access, no MFA required
The stolen token is injected into the attacker's browser. Microsoft sees a valid session and grants full access.
Why this fits this case: MFA was enabled, no malware on PC, and multiple logins from foreign locations — all consistent with token replay from different attacker-controlled IPs.
2. MFA Fatigue / Push Bombing
🟠 High likelihood — Second possibilityThe attacker already has the user's password (obtained via phishing or credential reuse from a data breach). They then repeatedly trigger MFA push notifications to the user's phone — sometimes dozens of times in rapid succession — until the user approves one either accidentally or out of frustration.
Key warning sign: Did the user report receiving multiple unexpected MFA prompts before the compromise? If yes, push bombing is confirmed. This method was used in the high-profile Uber breach of 2022 and remains one of the most common MFA bypass techniques in 2025–2026.
3. Session Token Theft from Android Device
🟡 Possible — Phone as the primary vectorIf the user had Microsoft 365 logged in via browser or native app on their Android phone, a malicious application or browser exploit could have silently stolen the session tokens stored on the device. Possible vectors include:
- Fake or repackaged apps downloaded outside the Google Play Store
- Accessibility abuse — apps that misuse Android's Accessibility API to read and exfiltrate tokens
- Malicious browser extensions on mobile Chrome
- Compromised WiFi session hijacking (rare on HTTPS but possible with certificate pinning bypasses)
Mobile devices are typically less monitored than corporate laptops — no EDR agent, no DNS filtering, no Mobile Device Management (MDM). They represent a growing blind spot in enterprise security.
4. OAuth App Consent Attack
🟡 Possible — Requires immediate checkThe user unknowingly granted access to a malicious third-party application via a "Login with Microsoft" button. Once consent is granted, the attacker accesses the mailbox, calendar, and files through Microsoft's own API — bypassing MFA entirely because the authentication already happened during consent.
Check immediately: Azure AD → Enterprise Applications → look for suspicious apps with permissions including Mail.Read, Files.ReadWrite, or offline_access.
5. SIM Swap / SMS OTP Interception
🟢 Less likely if using Microsoft Authenticator pushIf the account was using SMS-based MFA, an attacker who performed a SIM swap with the mobile carrier would receive all OTP codes. This is why SMS MFA is considered deprecated by NIST SP 800-63B and should be replaced with authenticator apps or hardware keys immediately.
6. Credential Reuse from Data Breach
🟢 Supporting factor — Not a standalone bypassIf the user reused a password from another breached service, the attacker would already have valid credentials. This alone does not bypass MFA — but it is the necessary first step for methods 1 and 2 above. Check haveibeenpwned.com for the affected email address.
Why Your Phone Is the Weakest Link
Even when a corporate laptop is fully protected — with EDR agents, DNS filtering, and web proxies — the user's personal smartphone often operates in a completely unmonitored environment. Consider what happens on a typical working day:
- Links arrive via WhatsApp from colleagues, suppliers, and unknown numbers
- SMS messages carry fake delivery notifications, bank alerts, and "verify your account" prompts
- Personal email on the phone contains phishing lures indistinguishable from legitimate notifications
- Apps downloaded outside official stores may carry hidden payloads
- Small screens make it harder to inspect URLs before clicking
Research from Microsoft's Threat Intelligence team confirms that the majority of real-world M365 compromises originate from mobile phishing, not desktop malware. Attackers have followed users to mobile — and most organisations' security tooling has not kept up.
What to Check Immediately After a Suspected Compromise
If you suspect your Microsoft 365 account has been compromised, act in this sequence. Time matters — attackers often set up persistence within minutes of gaining access.
| Where to Look | What to Check | Red Flag |
|---|---|---|
| Azure AD Sign-in Logs | Filter by user, look at IP addresses, locations, and MFA status | CRITICAL "MFA satisfied by token" with foreign IP = AiTM confirmed |
| Azure AD Audit Logs | Check for new inbox rules, forwarding rules, consent grants, added auth methods | CRITICAL Forwarding to external email = data exfiltration in progress |
| Enterprise Applications | Review third-party apps with Mail.Read, offline_access, Files.ReadWrite permissions | HIGH Unknown apps with broad permissions = OAuth consent attack |
| Mailbox Rules | Check for hidden inbox rules that move, delete, or redirect emails | HIGH Rules created around compromise time = BEC staging |
| Authentication Methods | Verify no new phone numbers or authenticator apps were added | HIGH New MFA method = attacker-established persistence |
| Sent Items & Deleted Items | Look for outbound phishing emails sent to contacts | HIGH Emails to contacts with links = attack spreading to your network |
- Revoke all active sessions in Azure AD (Global Admin → User → Revoke Sessions)
- Reset the compromised user's password immediately
- Force re-registration of MFA using a trusted, monitored device
- Remove any suspicious OAuth app permissions found in Enterprise Applications
- Delete any malicious inbox rules
- Notify affected contacts if outbound phishing emails were sent
How to Prevent MFA Bypass Attacks: A Practical Checklist
The following controls should be implemented in priority order. The first two are the most impactful and directly counter the attack methods described in this incident.
1 Deploy Phishing-Resistant MFA
Switch to FIDO2 hardware security keys (YubiKey, Feitian) or passkeys. These bind authentication cryptographically to the legitimate domain — an AiTM proxy page receives nothing it can replay. SMS and basic push approval should be disabled for all privileged accounts.
2 Enable Conditional Access Policies
In Microsoft Entra ID: block sign-ins from high-risk countries, require compliant managed devices, enforce sign-in risk and user risk policies, and block legacy authentication protocols (SMTP, IMAP, POP3) which bypass MFA entirely.
3 Enable Number Matching on Push MFA
Turn on number matching in Microsoft Authenticator. This requires users to type a number shown on the login screen into their phone — making push bombing attacks far less effective since the attacker cannot supply the matching number.
4 Deploy Microsoft Defender for Office 365
Enable Safe Links (real-time URL rewriting and checking) and Safe Attachments (sandboxed detonation of email attachments). Safe Links can block AiTM phishing URLs before the user ever clicks them.
5 Enrol Mobile Devices in MDM
Use Microsoft Intune to enrol and manage employee smartphones. This enables remote wipe, enforces screen lock, blocks sideloaded apps, and allows Conditional Access to require device compliance before granting M365 access.
6 Conduct Regular User Awareness Training
Train every user on three rules: (1) Never approve an unexpected MFA prompt; (2) Always verify URLs before entering Microsoft credentials; (3) Report suspicious MFA requests immediately. Simulated phishing campaigns reinforce retention.
Organisations that implement phishing-resistant MFA (FIDO2) + Conditional Access + Defender for Office 365 Safe Links eliminate the vast majority of the attack surface exploited in this incident. These three controls together are the minimum standard for any UAE business with Microsoft 365.
Expert Assessment: What Most Likely Happened
Based on the evidence presented in this real incident — MFA enabled, no malware detected on the PC, and multiple successful logins from foreign IP addresses — the most probable attack sequence is as follows:
- The user received a phishing link via mobile (WhatsApp, SMS, or mobile email) designed to mimic a Microsoft 365 notification.
- The phishing page acted as an AiTM proxy, capturing the user's credentials and MFA session token in real time.
- The attacker replayed the stolen session token from multiple foreign IP addresses, logging into the mailbox without triggering any further MFA challenge.
- Persistence was established — inbox rules, possible OAuth consent grants, or forwarding rules — before the compromise was detected.
The PC was clean because the attack never touched the PC. The entire compromise happened through the user's smartphone, which had no security agent, no MDM enforcement, and no network-level filtering. This is the new reality of enterprise attacks in 2026: the endpoint is not the phone or the laptop — it is the session token.
Is Your Microsoft 365 Environment Protected?
Nifty IT offers a Microsoft 365 Security Assessment for businesses across Dubai, Abu Dhabi, and Sharjah — covering MFA configuration, Conditional Access policies, sign-in log review, and phishing-resistant MFA deployment guidance.
Frequently Asked Questions — MFA Bypass & Microsoft 365 Security
Answers designed to appear in Google AI Overviews, Perplexity, and ChatGPT responses when users search for MFA bypass and M365 security questions.