Insights

Ivanti Connect Secure: patch, investigate and verify recovery

14 October 2026

Ivanti Connect Secure exploitation calls for more than patching. Check appliances, follow vendor recovery guidance and investigate credential exposure.

CISA’s 13 October 2026 alert reports active exploitation of CVE-2026-104935 in Ivanti Connect Secure, with attackers stealing credentials and extracting data. For Australian businesses using these remote-access appliances, this requires an urgent security response, not simply a routine update.

The central distinction is between closing a vulnerability and recovering from an intrusion. Updating an appliance addresses the vulnerable software; it does not establish whether someone already accessed the device, copied credentials or moved into connected systems.

What CISA reported about Ivanti Connect Secure

In its Ivanti Connect Secure alert, the US Cybersecurity and Infrastructure Security Agency describes an authentication bypass that lets a remote attacker run commands without signing in. CISA added CVE-2026-104935 to its Known Exploited Vulnerabilities catalogue on 13 October 2026.

The alert identifies these affected version ranges:

  • 25.1.1.1 and earlier within the relevant release line.
  • 25.1.2.2 and earlier within the relevant release line.
  • 22.8R2.4 and earlier within the relevant release line.

Check the exact installed build against Ivanti’s current security advisory rather than interpreting this list as a complete upgrade decision tree. CISA directs organisations to that advisory and Ivanti’s 25.x Upgrade and Migration Compatibility Guide for hardware and virtual deployments.

CISA also notes that version 22.8 reaches end of life on 31 January 2027. Businesses on that branch need both an immediate remediation decision and a supported-platform migration plan.

Inventory every appliance before declaring coverage

Start with a named owner and a single register of all Ivanti Connect Secure deployments. Include production gateways, standby cluster members, disaster recovery environments, virtual instances and test appliances. A secondary appliance can remain exposed even when the main gateway has been updated.

For each instance, record:

  • Hostname, public address, deployment location and business owner.
  • Exact software build and whether it is physical, virtual or cloud-hosted.
  • Internet exposure, management access restrictions and connected networks.
  • Authentication sources, privileged accounts and stored integration secrets.
  • Cluster membership, backup status and available log retention.
  • The person responsible for remediation and evidence of completion.

Compare the register with firewall rules, virtual infrastructure inventories and remote-access documentation. Ask service providers to confirm assets under their management rather than accepting a general statement that patching is complete.

Also identify who depends on each gateway. Finance staff, administrators and external support teams may require an alternative access method during containment or recovery. That alternative should be approved and restricted, not an improvised opening in the firewall.

Confirm the vendor’s remediation path

Do not guess the fixed release or assume the next version number is sufficient. Use the Ivanti advisory linked from CISA’s announcement to confirm the target build, supported upgrade sequence and any platform-specific requirements.

Before making changes, ask the responsible engineer to document:

  • Which advisory revision they are following.
  • The source and integrity checks for the installation image.
  • Compatibility requirements for the appliance and migration path.
  • Configuration backup and recovery arrangements.
  • The expected outage, validation steps and rollback limitations.

A rollback to vulnerable software is not a secure recovery outcome. If an upgrade cannot proceed safely, escalate the exposure decision to the incident owner and consider restricting or isolating the affected service while a supported path is arranged.

Keep the emergency change record concise but useful. It should show what changed, who approved it and how the installed version was verified afterwards. An accepted maintenance job is not the same as evidence that every cluster member is running the intended release.

Preserve evidence before disruptive recovery work

CISA says network and memory forensics are needed to investigate potential compromise. It also warns that Ivanti’s Integrity Checker Tool may miss exploitation of this vulnerability because attacker commands can run in memory.

That makes a clean integrity-check result insufficient to rule out an intrusion. Missing alerts or apparently normal operation are not proof either.

Before rebooting, resetting or rebuilding an Ivanti Connect Secure appliance, involve a qualified incident responder to decide what evidence can be collected safely. Memory evidence can disappear during a restart. Balance preservation against the need to stop ongoing attacker access; do not leave a dangerous exposure open simply to complete a perfect collection.

Useful investigation inputs may include:

  • Available appliance, administrator and authentication logs.
  • Firewall, DNS, proxy and network traffic records.
  • Identity-provider and directory sign-in events.
  • Endpoint alerts on systems reachable through the gateway.
  • Configuration changes, unexpected connections and unusual outbound transfers.

Record collection times and preserve copies outside the potentially affected environment. Establish an investigation window using exposure history and available evidence, not just the announcement date. Our first-24-hours incident response checklist provides a related framework for assigning roles and coordinating containment, although this alert does not itself establish a ransomware incident.

Treat credentials and downstream access as a separate workstream

Because CISA reports credential theft, the investigation should extend beyond the gateway. Map what identities and secrets the appliance stores, uses or could expose, then assess where those credentials would work elsewhere.

CISA’s described recovery actions include resetting passwords across domain, administrator and local accounts. They also include revoking and replacing stored certificates, changing stored API keys and resetting licence server credentials. Do not narrow that scope without checking the current vendor recovery instructions and the incident findings.

Sequence changes carefully. Rotating credentials while an attacker retains access can expose the replacements. Coordinate containment, appliance recovery and credential changes, with application owners checking dependencies to avoid preventable outages.

Check whether active sessions or tokens also need revocation under the relevant identity platform’s procedures. Review privileged sign-ins, newly created accounts, permission changes and access to sensitive file stores. Strong authentication remains important, but it is not a substitute for fixing a flaw that bypasses authentication.

Use the investigation to test network boundaries as well. Could the gateway reach management systems or server networks unnecessarily? The network segmentation checklist can help identify which connections should remain permitted after recovery.

Follow recovery procedures and define closure evidence

Where compromise is suspected, CISA points to Ivanti’s recovery procedures. The actions it describes include backing up the configuration, factory-resetting hardware and upgrading it, or rebuilding cloud and virtual instances and installing the current release. Configuration restoration and credential replacement follow within that recovery process.

These are vendor-directed recovery actions, not a reason to skip evidence preservation. Confirm the latest instructions and review the configuration being restored for unexpected users, settings or access rules. Treat configuration backups as sensitive material and restrict access to them.

Before closing the incident, require evidence that:

  • Every in-scope appliance has been accounted for and remediated.
  • Suspected-compromise recovery steps have been completed where required.
  • Credential, certificate and key changes have been documented.
  • Relevant identity and connected-system activity has been investigated.
  • Remote access works with the intended restrictions and monitoring.
  • Remaining evidence gaps and residual risks have an accountable owner.

If investigation suggests data was accessed or removed, involve the organisation’s privacy and legal advisers to assess notification obligations. Exploitation reports do not automatically establish a reportable breach at an individual business, but they warrant a documented assessment.

Review your cyber security response

The practical goal is a supported appliance, controlled access and an evidence-based recovery decision. A successful patch is one milestone, not the whole result.

Tech Engine can help you review your cyber security, remote-access exposure and managed IT controls. Call 1300 088 324 or email sales@techengine.au to discuss your next steps.

Want this applied to your business?

Request an AI Blueprint and we will map the processes worth automating first.

Request an AI Blueprint