Guard Policy CloudEnterprise Underwriting Intelligence
Underwriting

Seven Remote Access Mistakes That Get Cyber Insurance Applications Declined

Underwriters scan your perimeter before they read your answers. Seven remote-access gaps most often get a cyber quote declined or loaded. Each comes with the fix and the evidence that clears it.

By the Guard Policy Cloud editorial team · November 2026 · 6 min read

Leaving RDP Reachable From the Internet

Many cyber carriers run an external scan of your domains and IP ranges before an underwriter reads your application. Whatever it finds frames the conversation. In most markets, the remote access requirements for cyber insurance begin with one question: can an attacker reach a login prompt on your network from the open internet? An exposed Remote Desktop Protocol (RDP) port answers yes. Many carriers will decline on that alone, or quote only once it's closed.

RDP exposure matters because it's a well-known way in for ransomware crews, and scanning for it costs them almost nothing. Moving RDP to a non-standard port doesn't fix it. Scanners find the service anyway, and underwriters know that.

Advertisement

What to do instead: close inbound RDP at the firewall. Put remote desktop behind a VPN or gateway that requires MFA, or behind a remote desktop gateway or zero-trust broker. Evidence underwriters accept: a clean rescan from the carrier or broker, or a current report from a tool such as Shodan or Censys showing no RDP on your ranges, plus a firewall rule export.

Running VPN and Firewall Appliances Weeks Behind on Critical Patches

VPN concentrators and firewalls from the major vendors have repeatedly had critical flaws that attackers exploited quickly after disclosure. Underwriters look less at your general patching policy and more at one thing: was an edge device running a known-exploited flaw on the day they scanned you? A vulnerable firmware banner often brings a subjectivity, meaning patch within a set number of days or the quote lapses.

Edge devices need a faster cycle than your monthly routine. A reasonable internal standard for internet-facing appliances is a few days, roughly 72 hours, for flaws on the CISA Known Exploited Vulnerabilities list. Carriers outside the US reference that list too, alongside advisories from national CERTs such as the UK's NCSC, CERT-In and Singapore's CSA.

Evidence underwriters accept: a written edge-patching SLA, dated change records for your last two critical advisories, and the current firmware version of each appliance.

Enforcing MFA on Email but Not on Every Remote Path

Applications now ask whether MFA protects all remote access. The honest answer is often "mostly." The usual gaps: a legacy VPN profile for a few executives, a firewall or hypervisor admin console reachable remotely, a Citrix or remote-support portal set up years ago, and service accounts exempt from conditional access.

Tick "yes" while those exceptions exist and you take on misrepresentation risk at claim time. That's worse than a declined application. SMS codes are also increasingly viewed as weak, and some carriers now ask about phishing-resistant methods such as FIDO2 keys or number-matching push.

What to do instead: list every way into the network, enforce MFA on each, and document any exception you can't remove with compensating controls such as IP allow-listing and credential vaulting. Evidence: identity-provider policy screenshots, the exception register and a remote-access inventory.

Giving Vendors Shared, Always-On Accounts

Your MSP, payroll provider or equipment maker probably has remote access. If they connect with a shared login that never expires and skips MFA, that's a red flag underwriters see often in third-party vendor access reviews. One compromised vendor account can open a path into many clients at once.

The standard underwriters want: a named account for each vendor technician, MFA on every login, access granted only when needed and switched off after the session, and session logging. Privileged access tools such as BeyondTrust, CyberArk or Delinea handle this. For a mid-size company, budget roughly $3,000 to $25,000 a year depending on vendor and session counts. These are typical 2026 market ranges that vary by contract, so confirm with the vendor.

Evidence: a vendor access register, sample session logs and contract clauses requiring vendor MFA. Processor obligations under GDPR-style laws in Europe, and comparable rules in markets such as Brazil and India, can give you a contractual basis to demand these controls. Check the specifics with counsel.

Keeping End-of-Life Edge Devices in Service

A firewall or VPN appliance that no longer receives security updates can't pass the patching test, however carefully you run it. Underwriters can often identify the model and firmware from scan banners. A product past end-of-support commonly means a decline or a meaningful premium load until it's replaced.

Replacement is a budget decision, so plan it before renewal. A small-office next-generation firewall typically runs $1,500 to $6,000 including first-year subscriptions. A mid-size high-availability pair can run $15,000 to $60,000 or more. These are indicative 2026 ranges. Local pricing and import duties differ by country.

What to do instead: keep an inventory of edge devices with vendor end-of-support dates, and replace each one well before it lapses; six months is a sensible buffer. Evidence: the inventory itself, plus a purchase order or dated migration plan if a replacement is still in progress at renewal.

Treating the VPN as the Only Security Boundary

A VPN that drops every user onto a flat internal network means one stolen credential exposes everything. Underwriters increasingly ask about segmentation and least-privilege access, which is where ZTNA enters cyber insurance discussions. Zero-trust network access connects each user to specific applications rather than the whole network, and it can remove the exposed VPN appliance altogether.

Widely available options include Zscaler Private Access, Cloudflare Zero Trust, Netskope, Palo Alto Networks Prisma Access and Microsoft Entra Private Access. Typical 2026 pricing runs from about $5 to $18 per user per month depending on tier and volume. Confirm with vendors and resellers in your region.

For most 50- to 500-person companies running mainly cloud and web apps, ZTNA is a better fit than replacing an ageing VPN appliance, because one project fixes both the patching problem and the segmentation problem. Keep a VPN only for legacy protocols ZTNA can't broker.

Answering the Application From Memory Instead of a Scan

The last mistake is a process failure. Finance or the broker fills in the application from last year's answers, then the carrier's scan contradicts it. Even a small mismatch costs you credibility with the underwriter and can complicate a claim later. Run your own external scan two to three months before renewal so there's time to fix what it finds.

IssueTypical fix timeEvidence to send
Exposed RDPDaysClean external rescan
Unpatched edge deviceDays to 2 weeksChange records, firmware versions
MFA gaps2 to 6 weeksIdP policies, exception register
Shared vendor accounts1 to 3 monthsVendor register, session logs
End-of-life appliance1 to 4 monthsInventory, purchase order
Flat VPN access2 to 6 monthsZTNA rollout plan or segmentation diagram

Fix times are rough planning figures and depend on your environment.

Decision rule: if the scan finds RDP or an unpatched edge device, fix it before you submit. If a finding will take longer, send a dated remediation plan with the application. A cyber-specialist broker can tell you which subjectivities a given carrier accepts. Have a licensed professional in your jurisdiction review the coverage terms.

Frequently asked questions

Will one exposed RDP port really get my application declined?

With many carriers, yes, or the quote is issued only on condition that the port is closed before binding. Some markets will offer terms with a higher premium or a ransomware sublimit instead. The fastest route is to close the port and send a clean rescan.

Is a VPN with MFA enough, or do insurers require ZTNA?

Most carriers accept a VPN with MFA for every user, provided the appliance is patched and still supported. ZTNA isn't a universal requirement. It can improve how your risk is viewed and removes the edge-patching problem. Ask your broker whether a given carrier gives credit for it.

How do underwriters know what our remote access looks like?

Many carriers and brokers run passive external scans of your domains and IP ranges. These pick up open ports, service banners and firmware versions on internet-facing devices. They then compare the results with your application answers and may ask about any differences.

What should we ask our IT provider or MSP before renewal?

Ask for a full list of remote access paths, MFA status for each, edge-device firmware versions with end-of-support dates, and how the MSP's own technicians log in to your systems. Get written evidence rather than verbal assurances, because that's what the underwriter will ask for.

Written by the Guard Policy Cloud editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.