A 2026 report from cyber insurance provider At-Bay highlights a significant evolution in ransomware tactics, with threat actors moving away from phishing and focusing on the direct exploitation of core network infrastructure. The analysis, based on thousands of insurance claims, reveals that 73% of all ransomware incidents in 2025 began with the compromise of a Virtual Private Network (VPN). The Akira ransomware group was the primary catalyst for this trend, responsible for over 40% of claims and showing a strong preference for exploiting SonicWall VPN appliances. The report underscores that traditional defenses are struggling to keep pace; 60% of Akira's victims had an Endpoint Detection and Response (EDR) solution deployed, yet were still compromised, pointing to the critical need for continuous monitoring and response capabilities.
The At-Bay report signals a strategic shift in the ransomware ecosystem. Instead of relying on user interaction (e.g., clicking a malicious link), ransomware groups like Akira are adopting a more direct, infrastructure-led approach. This method is more efficient, scalable, and often bypasses user-centric security controls.
This data indicates that ransomware has become a game of exploiting unpatched, internet-facing infrastructure. Groups like Akira systematically scan the internet for vulnerable devices and exploit them for initial access.
Attacks targeting VPNs typically fall into two categories:
Once inside, Akira and similar groups follow a standard ransomware playbook: escalate privileges, move laterally, exfiltrate sensitive data for double extortion, and finally, deploy the ransomware payload to encrypt systems.
The report's finding that 60% of victims had EDR is significant. It suggests that attackers are either using techniques to bypass or disable EDR, or that the EDR alerts were not acted upon quickly enough. This is where the value of a 24/7 Managed Detection and Response (MDR) service becomes apparent, as it provides the human element to investigate and respond to alerts around the clock.
T1190 - Exploit Public-Facing Application, T1133 - External Remote ServicesT1059.001 - PowerShellT1078 - Valid AccountsT1021.002 - SMB/Windows Admin SharesT1041 - Exfiltration Over C2 ChannelT1486 - Data Encrypted for ImpactThe business impact of these infrastructure-led attacks is severe. The average cost for smaller businesses reached $422,000, a figure that can be existential. The shift to targeting core infrastructure means that attacks can be more disruptive, potentially taking down entire networks rather than just individual machines. The high ransom demands from groups like Akira put immense financial pressure on victims. Furthermore, the double extortion tactic of exfiltrating data before encryption adds the costs of data breach notification, credit monitoring, and reputational damage to the recovery equation.
No specific Indicators of Compromise were provided in the report summary.
Security teams should focus on hunting for signs of VPN compromise:
powershell.exe, psexec.exe, wmic.exe/api/ssl-vpn/D3FEND Techniques:
D3-MFA: Multi-factor Authentication: The most critical defense for VPNs.D3-SU: Software Update: Essential for closing the vulnerabilities that groups like Akira exploit.D3-NI: Network Isolation: Use network segmentation to contain the blast radius of a VPN compromise.Enforcing MFA on all VPN connections is the most critical defense against credential-based attacks.
Mapped D3FEND Techniques:
Aggressively patch internet-facing infrastructure like SonicWall VPNs to close known vulnerabilities.
Mapped D3FEND Techniques:
Isolate VPN user traffic into a restricted zone to limit the blast radius of a compromised session.
Mapped D3FEND Techniques:
The At-Bay report confirms that VPNs are the primary entry point for ransomware. The single most effective countermeasure is to enforce phishing-resistant Multi-factor Authentication (MFA) on all remote access solutions, including SonicWall VPNs. This is not optional. Standard username/password combinations are insufficient against credential stuffing and other password-based attacks. Organizations should prioritize the rollout of MFA using methods like FIDO2/WebAuthn hardware tokens, which are resistant to phishing, or at a minimum, authenticator apps. This single control would have prevented a large percentage of the credential-based VPN compromises that lead to ransomware events. The implementation should be comprehensive, covering not just standard users but also administrative accounts, service accounts, and third-party contractors with VPN access.
Assuming a VPN compromise will eventually happen, Network Isolation (or segmentation) is the next critical defense. Instead of granting VPN users broad access to the entire corporate network, their sessions should terminate in a highly restricted 'DMZ-like' zone. From this zone, firewall rules should explicitly whitelist access only to the specific servers and applications that the user role requires. All other traffic, especially broad network scanning and access to critical infrastructure like Domain Controllers or backup servers, should be denied by default. This containment strategy severely limits an attacker's ability to perform reconnaissance and move laterally after gaining an initial foothold via a compromised VPN account. It turns a potential network-wide disaster into a contained security incident.
To detect compromised VPN credentials, security teams should implement User Geolocation Logon Pattern Analysis. This involves ingesting all VPN authentication logs into a SIEM or security analytics platform and enriching them with geolocation data. The system should then build a baseline of normal login locations for each user. High-fidelity alerts should be configured for anomalies, such as: 1) 'Impossible travel,' where a user logs in from two geographically distant locations in a time frame that would be impossible to travel between. 2) Logins from countries where the organization has no employees or business operations. 3) A single IP address being used to attempt logins for multiple different user accounts. These patterns are strong indicators of credential stuffing or account sharing and should trigger an automated response, such as forcing a password reset and MFA re-enrollment.

Cybersecurity professional with over 10 years of specialized experience in security operations, threat intelligence, incident response, and security automation. Expertise spans SOAR/XSOAR orchestration, threat intelligence platforms, SIEM/UEBA analytics, and building cyber fusion centers. Background includes technical enablement, solution architecture for enterprise and government clients, and implementing security automation workflows across IR, TIP, and SOC use cases.
Help others stay informed about cybersecurity threats