Your Tenant Answers Half the Application Before You Do
Most cyber applications ask the same handful of control questions: MFA on email and remote access, MFA on privileged accounts, email filtering, endpoint detection and response (EDR), logging, and backups. At a company that runs on Microsoft 365, the answers to four of those sit in the tenant. If the honest answer is "partially" or "for most users," underwriters will load the premium, attach an exclusion or decline to quote.
This checklist puts Microsoft 365 security for cyber insurance work in the order that removes the most underwriting risk first. Each step names the setting, the application question it answers, the license it needs, and the screenshot or report a broker can attach to the submission. Start eight to ten weeks before renewal. Conditional access changes pushed into enforcement too fast tend to lock out executives, and then they get rolled back.
Questionnaires differ by carrier and market. London, continental Europe, India, the Gulf, Singapore and Latin America all phrase them differently, so confirm the exact wording with your broker before you build evidence.
Looking to streamline vendor RFP selection?
Access pre-negotiated tier discounts through our independent institutional network.
What Business Premium, E3 and E5 Can Actually Prove
Your license decides which controls you can enforce and which you can only describe. Prices below are indicative typical 2026 list ranges per user per month on annual commitment. Local-currency pricing, reseller discounts and enterprise agreements vary, so confirm with your Microsoft partner. Plan contents change too; verify against Microsoft's current licensing documentation.
| Capability | Business Premium (roughly $20–25, 300-user cap) | Microsoft 365 E3 (roughly $35–40) | Microsoft 365 E5 (roughly $55–60) |
|---|---|---|---|
| Conditional access | Yes (Entra ID P1) | Yes (Entra ID P1) | Yes, plus risk-based policies (P2) |
| Privileged Identity Management | No (P2 add-on) | No (P2 add-on) | Yes |
| EDR | Defender for Business | Defender for Endpoint P1 (limited EDR) | Defender for Endpoint P2 |
| Email threat protection | Defender for Office 365 P1 | Exchange Online Protection only | Defender for Office 365 P2 |
| Audit log retention | 180 days (Standard) | 180 days (Standard) | 1 year (Premium) |
Note the E3 gap. Many mid-size firms assume E3 is "more secure" than Business Premium. For underwriting purposes it often isn't: E3 lacks Defender for Office 365 and has only P1-level endpoint protection. Under 300 seats, Business Premium is usually the stronger evidence position for the money. Above 300, E3 plus a Defender for Office 365 add-on and a full EDR, either Defender for Endpoint P2 or a third-party tool such as CrowdStrike or SentinelOne, closes most of the gap without moving to E5.
Steps 1–3: Lock Down Sign-In With Conditional Access
Step 1: Create two break-glass accounts. Use cloud-only Global Admin accounts with long passwords or FIDO2 keys, excluded from conditional access policies and monitored with a sign-in alert. Without them, the next steps carry real lockout risk.
Step 2: Require MFA for all users on all cloud apps. Build conditional access policies in report-only mode first. Review the sign-in logs for two weeks, then enforce. This answers "Is MFA enforced for all remote access to email?" Underwriters generally accept a screenshot of the policy list showing On (not Report-only), the assignment set to All users with only break-glass exclusions, and the authentication methods registration report showing near-complete coverage. On tenants without Entra ID P1, turn on Security Defaults instead and screenshot the toggle.
Step 3: Block legacy authentication. Microsoft has retired Basic auth for most Exchange Online protocols, but scanners, multifunction printers and old line-of-business apps still turn up on SMTP AUTH or other legacy clients. Filter sign-in logs by client app to find them first, then create a conditional access policy that blocks "Exchange ActiveSync clients" and "Other clients." Legacy protocols skip MFA entirely, and underwriters know it. Evidence is the policy screenshot plus a sign-in log export showing no successful legacy sign-ins over 30 days.
Steps 4–5: Shrink and Protect Privileged Access
Step 4: Cut Global Admins to between two and four. Tenants often carry far more than they need, including former IT contractors. Move people into scoped roles such as Exchange Administrator, User Administrator or Helpdesk Administrator. This answers "How many accounts have domain or tenant administrator rights?" Evidence is an export of role assignments from Entra ID.
Step 5: Require phishing-resistant MFA for admins. Use a conditional access policy with the built-in "Phishing-resistant MFA" authentication strength, targeted at directory roles and met with FIDO2 keys (typically tens of dollars each) or Windows Hello for Business. On E5, or with the Entra ID P2 add-on, add Privileged Identity Management so admin rights are activated just in time instead of standing. Many carriers now ask specifically about privileged access management, and PIM activation history is among the cleanest evidence you can give them.
Looking to streamline vendor RFP selection?
Access pre-negotiated tier discounts through our independent institutional network.
Steps 6–7: Close the Email Gaps Behind Most Claims
Business email compromise and funds-transfer fraud account for a large share of cyber claims, which is why email questions get the closest reading.
Step 6: Turn on preset security policies and publish DMARC. In Business Premium or E5, apply the Standard or Strict preset policies in Defender for Office 365 so Safe Links and Safe Attachments cover every mailbox. On E3, either buy the add-on or document your third-party gateway (Proofpoint, Mimecast and similar). For every sending domain, publish SPF, enable DKIM signing in the Defender portal, and move DMARC from p=none toward p=quarantine or p=reject. A p=none record shows monitoring has started, not that spoofed mail is being stopped. Evidence is the preset policy screenshot plus a DNS lookup of all three records.
Step 7: Block automatic external forwarding. Set the outbound spam policy's auto-forwarding control to Off, and allow approved exceptions by named mailbox. Attackers use silent forwarding rules to watch invoice threads for weeks. This answers BEC and wire-fraud control questions, and some carriers ask for it by name.
Steps 8–9: Prove You Can Investigate and Detect
Step 8: Confirm mailbox audit logging and the unified audit log. Mailbox auditing is on by default for most licensed mailboxes, but older tenants, or those with per-mailbox bypasses, sometimes have gaps. Run Get-OrganizationConfig in Exchange Online PowerShell and confirm AuditDisabled is False. Confirm unified audit logging is on in Purview too. Audit (Standard) keeps records for a limited period, currently 180 days. E5's Audit (Premium) keeps them longer and records MailItemsAccessed events, which forensic firms use to establish which messages an attacker actually read. That bears on breach-notification decisions under laws such as GDPR, India's DPDP Act and Singapore's PDPA. On Business Premium or E3, consider exporting logs to a SIEM for longer retention.
Step 9: Onboard every endpoint to EDR. Onboard devices to Defender through Intune, with tamper protection on and EDR in block mode where available. This answers "Do you deploy EDR on all endpoints and servers?" Underwriters accept a device inventory export showing onboarding status. Servers are the usual gap, because Defender for Servers is licensed separately through Defender for Cloud.
Step 10: Turn Microsoft Secure Score Into an Evidence Pack
Microsoft Secure Score is available in every tenant, and some underwriters and MGAs ask for it outright. Don't chase a number. Marking items "resolved through third party" or "risk accepted" without justification inflates the score and undermines your credibility if an adjuster later compares it with reality. Use it as a gap finder, export the recommendations history, and bundle it with the evidence below.
| Application question | Setting | Evidence to attach |
|---|---|---|
| MFA on remote email access | Conditional access, all users | Policy list screenshot plus registration report |
| Legacy protocols disabled | Block legacy auth policy | 30-day sign-in log filter |
| Privileged account controls | Scoped roles, phishing-resistant MFA, PIM | Role assignment export |
| Email filtering and spoofing | Preset policies, SPF/DKIM/DMARC | Policy screenshot plus DNS records |
| Logging | Unified and mailbox audit logging | PowerShell output plus retention setting |
| EDR coverage | Defender onboarding | Device inventory export |
Date every screenshot and include the tenant name. An exhibit that could come from any tenant proves little about yours.
An Eight-Week Sequence and the Red Flags That Stall Quotes
A workable sequence for a 50 to 500-person company:
- Weeks 1–2: break-glass accounts, admin role cleanup, conditional access in report-only mode, Secure Score baseline.
- Weeks 3–4: enforce MFA and the legacy authentication block, deploy keys to admins.
- Weeks 5–6: email presets, forwarding block, DMARC moving to quarantine.
- Weeks 7–8: audit checks, EDR onboarding of remaining devices, assemble the evidence pack.
Red flags that slow underwriting or change terms:
- MFA exclusions for executives, shared mailboxes with sign-in enabled, or entire service-account groups.
- Answering "yes" to MFA while policies are still in report-only mode. A misstatement can jeopardize coverage.
- More than a handful of Global Admins, or admin accounts also used for daily email.
- DMARC stuck at p=none for a year or more.
- EDR on laptops but not on servers.
Decision rule: if you can't produce dated evidence for a control, answer "in progress" with a completion date, and have your broker present it that way. Disclosed gaps get priced. Undisclosed ones get litigated. For policy wording, warranties and how your carrier treats misstatements, consult your broker or a licensed insurance advisor in your jurisdiction.