On 8 June 2026 Check Point published a hotfix for CVE-2026-50751, an authentication bypass in the IKEv1 key exchange of its Security Gateways. By then the flaw had been exploited for a month: Check Point dates the earliest observed exploitation to 7 May 2026, and in at least one case it confirmed post-compromise activity linked to an affiliate of the Qilin ransomware operation. Looking back at the case from November, it is a clean illustration of a problem every IT manager faces: for weeks there was nothing to patch, and the only control that could still work was noticing that a VPN session did not belong to the person whose name it carried. This article reconstructs the case from the vendor advisory and the CISA catalog, then turns it into concrete detection and hardening steps. For the broader picture of how ransomware groups choose their way in, see our ransomware intelligence guide.
Key Takeaways
- CVE-2026-50751 opened a remote-access VPN session without a valid user password on Check Point gateways configured with the deprecated IKEv1 protocol. CVSS 9.3.
- Exploitation started on 7 May 2026, the fix arrived on 8 June: about a month in which patching was not an option.
- CISA added it to the KEV catalog on 8 June with known ransomware use, and Check Point linked one intrusion to a Qilin affiliate (medium confidence).
- A VPN session is not an attack yet. The attacker still needs to move inside the network, and that is where detection has its chance.
- Disable IKEv1 for remote access where you can, apply the hotfix, and review logs back to 7 May if you ran an affected configuration.
What Happened, Date by Date
The timeline below comes from Check Point's advisory and the CISA Known Exploited Vulnerabilities (KEV) catalog.
| Date (2026) | Event | Source |
|---|---|---|
| 7 May | Earliest exploitation observed by Check Point | Check Point |
| 13 May | Palo Alto Networks publishes CVE-2026-0257, a GlobalProtect authentication bypass | Palo Alto Networks |
| 29 May | CVE-2026-0257 enters the KEV catalog, ransomware use "Known" | CISA KEV |
| 4 June | Check Point Research starts investigating suspicious activity | Check Point |
| 8 June | Advisory and hotfix for CVE-2026-50751; same-day KEV entry, ransomware use "Known", due date 11 June | Check Point, CISA KEV |
| 9 to 11 June | Further indicators of compromise added to the advisory | Check Point |
According to the Check Point advisory, a logic flaw in certificate validation allowed an attacker to establish a VPN session without a valid password. The affected products are Remote Access VPN, Mobile Access and Spark firewalls, across versions from R80.20.X up to R82.10, several of which are already end of support. Only deployments that use IKEv1 are exposed. Check Point described the scope as "a few dozen targeted organizations globally" and assessed, with medium confidence, a financially motivated actor linked to Qilin. A second flaw disclosed the same day, CVE-2026-50752 (CVSS 7.4), affects site-to-site IKEv1 certificate validation and was not seen in the wild.
Two details from the advisory matter for defenders. First, the attackers used virtual private servers geolocated in the same country as some of their victims, which defeats the simplest geography rule. Second, Check Point states that "additional post-authentication activity is required to access internal resources or escalate privileges". The VPN session was the door, not the attack.
A Month Without a Patch: Why Detection Is the Control That Remains
Most vulnerability programmes are built around a simple loop: a fix is published, you apply it within a target time, you verify. That loop assumes the fix exists. Between 7 May and 8 June 2026 it did not, and an organisation running an exposed IKEv1 configuration had three options only: reduce exposure (turn off IKEv1 remote access, if it knew it was using it), notice the abuse, or be lucky.
This is not a rare situation for edge devices. VPN concentrators and firewalls sit on the internet by design, they authenticate strangers by design, and an authentication bypass on them turns the most trusted network segment, the remote-access pool, into the attacker's starting point. The dwell time that follows decides the outcome. An attacker who logs in on Tuesday and starts encrypting on Friday has given you three days. An organisation that reads its VPN and identity logs only when something breaks will not use them.
Not Just Check Point: the Spring of VPN Bypasses
CVE-2026-50751 was not isolated. Two weeks earlier, on 29 May, CISA had added CVE-2026-0257 to the KEV catalog, also with known ransomware use. It is an authentication bypass in the GlobalProtect portal and gateway of PAN-OS that lets an attacker establish an unauthorised VPN connection when authentication override cookies are enabled with a particular certificate configuration. Palo Alto Networks reported "limited exploit attempts on unpatched PAN-OS devices without mitigations applied" and offered two workarounds: a dedicated certificate for authentication override cookies, or disabling authentication override.
Check Point added that the infrastructure behind CVE-2026-50751 appeared to be exploiting VPN vulnerabilities from Palo Alto Networks, Fortinet and F5 as well. The lesson is not about one vendor. Remote-access gateways from every major vendor were targeted in the same season, and a defence that depends on one vendor's patch speed is fragile.
The Italian Context
Italy saw the same pattern. ACN's Operational Summary for August 2026, as reported by ICT Security Magazine, counted 49 ransomware claims against Italian entities that month, the highest in its twelve-month window, with Qilin and The Gentlemen each behind 23% of them. The summary recalls CSIRT Italia bulletin BL01/260528 of 28 May, on Qilin campaigns that gained initial access through perimeter devices (Ivanti MDM and Fortinet vulnerabilities) and through VPN credentials obtained by brute force or bought from initial access brokers. The CSIRT named the lack of multi-factor authentication as the main structural weakness.
Different CVEs, same entry point: the remote-access edge, followed by valid-looking sessions.
What a VPN Login Gone Wrong Looks Like
When the authentication itself is bypassed, the login event can look normal. The signals are in the context around it and in what happens next. In MITRE ATT&CK terms you are looking for External Remote Services (T1133) and Valid Accounts (T1078), followed by discovery and lateral movement.
| Signal | Where you see it | Why it matters |
|---|---|---|
| VPN session with no matching MFA or identity-provider event | VPN gateway logs versus IdP logs | A bypass skips the normal authentication chain, so the usual second event is missing |
| Source IP in a hosting provider rather than a consumer ISP | VPN logs enriched with ASN data | Check Point saw VPS infrastructure; country alone is not enough |
| Unusual hour, client version or device for that user | VPN and endpoint logs | Real users have habits; a bypass session rarely matches them |
| Connections from the VPN pool to many internal hosts in minutes | Firewall, NetFlow, EDR | Internal discovery is the first job after access |
| LDAP and Active Directory enumeration, new SMB or RDP sessions from VPN addresses | Domain controller and EDR telemetry | Lateral movement toward servers and backups |
| Outbound downloads of executables or ELF files from unknown hosts | Proxy, firewall, EDR | Check Point observed attempts to download malicious ELF files |
None of these signals is conclusive alone. Together, read in sequence, they tell a story: a session that skipped MFA, from a VPS, at 03:00, followed by a scan of the server subnet. That correlation is the work a security operations centre does, and it is exactly the work a small IT team cannot do at night.
What a 24/7 Service Changes
A managed detection service does not make an unpatched gateway safe, and nobody can promise it would have stopped a specific intrusion. What it changes is who is looking, and when. Orizon Oversight is built on the CYREBRO platform, whose integrations list counts 416+ technologies, including firewalls, EDR and identity providers. Events from those sources are normalised and correlated with rules mapped to MITRE ATT&CK, and analysts on duty 24/7 confirm, discard or dig deeper before anything reaches you. A validated detection arrives as an investigation with a timeline, evidence, affected systems and recommendations. Containment on integrated technologies such as the EDR follows what you agreed in advance.
For the VPN scenario, that means the gateway logs, the identity provider and the EDR are read together, so "session without MFA" and "scan from the VPN pool" become one story rather than two unrelated lines in two consoles. If you are weighing service models, our comparison of MDR, MSSP and SIEM explains the differences.
Hardening the Remote-Access Edge
For Check Point gateways
- Apply the hotfix on every affected Security Gateway and Spark firewall. Versions already end of support are a separate risk: plan their replacement.
- Find out whether IKEv1 is enabled for remote access. Many deployments keep it for legacy clients nobody remembers. If no client needs it, turn it off; Check Point describes alternative configuration mitigations in note sk185033.
- Require machine certificates in addition to user authentication where your client fleet allows it. According to coverage of the advisory by The Next Web, exploitation required a gateway that did not demand a machine certificate.
- Review logs back to 7 May 2026, the earliest exploitation date Check Point gives, and compare them with the published indicators of compromise.
For every VPN, whatever the vendor
- Put MFA on every remote-access path, and alert when a session appears without its MFA event.
- Treat the VPN pool as untrusted: restrict which internal subnets and ports it can reach, and keep management interfaces and domain controllers off-limits by default.
- Keep VPN, firewall and identity logs for long enough to investigate backwards. When the next zero-day is disclosed, the question will be "since when?".
- Track the KEV catalog for your edge devices and act on entries with known ransomware use first. Our article on prioritising vulnerabilities with KEV, EPSS and SSVC explains how.
- Check what is exposed. A VPN portal you forgot about is still a VPN portal. Knowing your external attack surface is the precondition for every point above.
Checklist: the Next Edge Zero-Day
- Keep an inventory of every remote-access gateway, its version, its support status and the protocols it accepts.
- Subscribe to your vendors' advisories and to the CISA KEV feed, and assign a named owner for edge devices.
- Define in advance who can switch off a protocol or a portal in an emergency, and how quickly.
- Correlate VPN, identity and endpoint logs, and make sure someone reads the alerts outside office hours.
- Rehearse the scenario "valid VPN session, attacker behind it" in your incident response plan.
- After every fix, look for signs of compromise from the first exploitation date, not from the patch date.
If your team cannot watch VPN, identity and endpoint signals around the clock, that is the gap Oversight is designed to close, with analysts on duty 24/7 and containment that follows the rules you set. You can see how the managed service works, and what you receive when a detection is confirmed, on the managed security page.