Search for "Microsoft 365 security misconfiguration statistics" and you'll find plenty of vendor pages quoting big round numbers with no source. We'd rather give you something you can actually act on: the misconfigurations that genuinely recur across M365 tenants, why they matter, and exactly how to check your own environment for each one.
Where we cite a benchmark or percentage below, it comes from a named public source — the CIS Microsoft 365 Foundations Benchmark, CISA guidance, or Microsoft's own documentation — not from invented data. When we've completed a large enough body of real audits to publish anonymized aggregate findings honestly, we'll do that in a follow-up and label it clearly.
Why misconfiguration beats "hacking" as a root cause
Most M365 compromises don't start with a clever exploit. They start with a default that was never changed, a legacy protocol that was never turned off, or an admin account that was never given MFA. Microsoft has stated publicly that enabling MFA blocks the overwhelming majority of account-compromise attacks — and yet a large share of tenants still have gaps in exactly that control.
That's the frustrating and hopeful part: the highest-impact fixes are usually configuration changes, not new products. The categories below are the ones that come up again and again when you walk a tenant against the CIS benchmark.
1. Multi-factor authentication gaps
MFA is the single highest-leverage control in M365, and it's still the most common place to find holes. The typical failure modes:
- Admins without MFA. Even when end users have MFA, privileged roles (Global Admin, Exchange Admin, etc.) are sometimes exempted "for convenience."
- Partial coverage via per-user MFA. Legacy per-user MFA settings get left in a mixed state instead of being enforced consistently through Conditional Access or security defaults.
- No enforcement for external/guest identities.
How to check: In the Microsoft Entra admin center, review Conditional Access policies (or confirm security defaults are enabled on smaller tenants). Cross-reference the list of privileged role holders against your MFA registration report. The CIS benchmark explicitly recommends MFA for all users and especially all administrative roles.
2. Legacy authentication still enabled
Legacy auth protocols — IMAP, POP, SMTP AUTH, and older Exchange ActiveSync flows — bypass modern authentication, which means they bypass MFA and most Conditional Access. Attackers love them precisely because a stolen password alone is enough.
Microsoft has been deprecating basic authentication for years, but tenants that migrated gradually, or that kept a legacy app or scanner alive, often still have these pathways open.
How to check: Look at sign-in logs filtered for legacy authentication clients, and review the "Block legacy authentication" Conditional Access policy. SMTP AUTH can also be disabled at the organization level and re-enabled only per-mailbox where a genuine line-of-business need exists.
3. Over-privileged and standing admin access
Two related problems show up constantly:
- Too many Global Admins. Microsoft recommends keeping Global Administrator assignments to a small number (commonly cited as fewer than five), with everyone else scoped to the least-privilege role they actually need. In practice many tenants have far more.
- Standing (always-on) privilege instead of just-in-time elevation through Privileged Identity Management (PIM), where licensing allows.
Over-privileged accounts turn any single compromise into a tenant-wide one.
How to check: Enumerate role assignments in Entra ID. Count Global Admins, confirm each one is still justified, ensure break-glass accounts are documented and MFA-protected, and move eligible tenants toward PIM for just-in-time elevation.
4. Mailbox forwarding and malicious inbox rules
This one is subtle and dangerous. A common post-compromise technique — and a sign a breach already happened — is an auto-forwarding rule or inbox rule that silently copies or deletes messages (for example, hiding payment-fraud replies). External auto-forwarding is a frequent finding because it's easy to set up and easy to miss.
How to check: Review the organization-wide outbound spam policy to confirm automatic external forwarding is blocked by default, and audit mailboxes for forwarding rules and suspicious client-side inbox rules. The CIS benchmark calls out disabling external forwarding as a hardening step.
5. Unrestricted user consent to third-party apps
By default, users can sometimes grant OAuth consent to third-party applications that request access to their data. Attackers use illicit consent grant attacks to get persistent access without ever needing a password. If app consent is left wide open, a convincing phishing page is all it takes.
How to check: In Entra ID enterprise application settings, review the user consent policy. The hardened posture is to restrict user consent to verified publishers and admin-approved applications only, with a clear request-and-review path for anything else.
The short version
None of these are exotic. Every one of them is a setting that already exists in Microsoft 365 and Entra ID — the gap is almost always "never got changed from the default" rather than "actively misconfigured." That's good news: fixing them is usually a policy change and a short validation window, not a project.
The harder part is finding them before an attacker does. A one-time review catches today's gaps; tenants drift, roles change, and new apps get consented to every month.
Run an audit to check your tenant against the CIS Microsoft 365 Benchmark and catch these before they're the reason for an incident report.