Skip to content
cyber-security.eu

Microsoft 365 security

Microsoft 365 is the central workspace and identity ecosystem for many organisations. This page outlines the key security building blocks without admin step-by-step instructions.

Who this page is for

IT leaders, Microsoft 365 administrators, security leaders and auditors making an honest assessment of a tenant.

Security relevant building blocks

Identities and [MFA](/en/mfa), conditional access, role model, mail security with anti-phishing and anti-spoofing, audit logs, Defender components for endpoint, identity, mail and cloud apps, and app permissions and consent.

Typical risks

[Phishing](/en/phishing) and consent phishing target identities directly.

Compromised mailboxes with forwarding rules are often noticed late.

Weak admin accounts remain a primary attacker goal.

Missing conditional access allows sign-ins from arbitrary countries and devices.

Unused audit logs make incident analysis hard.

Sensible measures

Phishing-resistant MFA on all admins, strict conditional access for privileged roles, separate admin identities, anti-phishing and anti-spoofing in Exchange Online, active audit logs with adequate retention, Defender components actually reviewed, and a consent strategy for third-party apps. Concrete configuration belongs in specialised admin documentation.

Scenario

An organisation spots a business email compromise because a forwarding rule to an external address appeared. Audit logs and Defender alerts make the case visible, conditional access limits the damage.

Checklist

  • MFA for all accounts, phishing-resistant for admins
  • Separate admin identities without mailbox
  • Conditional access with risk-aware policies
  • Anti-phishing and anti-spoofing active
  • Audit logs active and reviewed
  • Consent strategy for third-party apps
  • Defender components reviewed, not just enabled

Frequently asked questions

+Are default settings enough?

Defaults are a start but do not replace deliberate hardening.

+What is consent phishing?

Attackers request permissions via a malicious app. Approving it grants access without revealing a password.

Related topics