The National Cyber Security Centre is pressing UK small and medium enterprises to close internet-facing Remote Desktop Protocol ports after another month of scanning data showed cloud servers still advertising administrative access to the public internet.

Security teams treat exposed TCP 3389 and alternate RDP ports as low-hanging fruit for ransomware crews and credential-stuffing bots. NCSC guidance published over the summer warns that cloud defaults often publish management interfaces directly, widening the attack surface compared with on-premises networks kept behind VPNs.

Why RDP remains a target

Remote Desktop Services let administrators control Windows servers graphically, but when ports are open to the world they become login walls attackers probe continuously. Microsoft’s security team has noted that Shodan-style internet scans routinely find millions of reachable RDP endpoints, many using weak or reused passwords.

NCSC’s July blog on interactive administration in the cloud describes attackers exploiting misconfigured interfaces and legacy protocol flaws to gain footholds, then move laterally toward payroll, customer databases or operational technology networks.

What NCSC recommends instead

Organisations should place administration behind VPNs, jump hosts or managed remote-access services with multi-factor authentication. Privileged access workstations, patching within days of critical fixes, and logging every admin session are baseline expectations in NCSC’s secure administration collection.

For firms that migrated servers quickly during hybrid-working rollouts, the fix is often a firewall rule audit: security groups labelled “Remote Desktop” or “Terminal Services” may still allow inbound traffic from 0.0.0.0/0 even after staff return to offices.

Broader exposure risks

A separate NCSC alert this year highlighted internet-visible operational technology and edge devices, urging organisations not to assume OT is air-gapped without verification. While Sunday’s reminder focuses on SME Windows estates, the same discovery discipline applies to NAS appliances, VPN concentrators and unpatched firewalls.

Practical steps for Monday

IT contractors serving UK retailers and professional-services firms should run external scans against client IP ranges, register for NCSC Early Warning where eligible, and document who approved any remaining admin exposure. Insurance underwriters increasingly ask for evidence that RDP is not publicly reachable before renewing cyber policies.

Managed service providers

MSPs serving accountancy and legal firms often inherit legacy RDP rules when onboarding clients. NCSC recommends contractual clauses requiring external scans quarterly and immediate closure of admin ports unless customers sign risk acceptance forms.

Cyber insurers have begun asking whether clients use privileged access management; bare RDP with password-only auth increasingly fails underwriting questionnaires.

Cloud migration hangover

Lift-and-shift projects that copied on-prem images to IaaS frequently imported insecure remote-admin habits. NCSC’s cloud guidance urges administration proxies even when VPNs feel inconvenient for small IT teams working from home.

Reporting culture

Near-miss reporting for admin interfaces should feed into board packs alongside ransomware drills. NCSC’s disruptive-activity alert noted that attackers probe edge devices and OT networks with the same scanning infrastructure used against Windows RDP, so discovery tools should cover both IT and operational estates.

Supply chain

Retail franchises sharing IT support often inherit one exposed RDP server serving tills and back-office payroll; segmenting those functions onto separate VLANs reduces blast radius when credentials leak.

NCSC’s Early Warning service emails technical contacts when external scans match registered IP ranges—free for many UK organisations and underused by firms below 250 employees.

Local resilience forums can escalate repeated scan findings to regional police units when critical infrastructure providers are affected.