Breach Watch: Cisco Email Gateway Flaw Exploited

Breach Watch: Cisco Email Gateway Flaw Exploited
CISO PLATFORM · BREACH INTELLIGENCE
BREACH WATCH
Daily Breach Intelligence for Security Leaders · October 14, 2026

TL;DR for CISOs: Cisco confirmed attacks on a critical Secure Email Gateway flaw, Palo Alto Networks fixed an exploited PAN-OS management interface zero-day, Microsoft and Health-ISAC disrupted phishing kits tied to more than one billion messages, and attackers backdoored Nextcloud servers through the Collabora CODE app.

BOTTOM LINE FOR CISOS
  1. Gateways, firewall management planes and self-hosted collaboration servers are today's entry points. Know which are internet-reachable and who can reach their admin interfaces.
  2. A shipped fix is not a closed risk. Cisco and Collabora both fixed their flaws in September, and both were exploited into October. Track the time from patch release to deployment.
  3. Phishing is now sold as a service built on real cloud tenants. A takedown helps, but identity controls and tenant monitoring still do the work.

Lead story: Cisco Secure Email Gateway flaw CVE-2026-76519 is exploited for root access

CRITICAL · EMAIL SECURITY / EDGE
Key facts
Flaw: CVE-2026-76519, CVSS 9.8, in Cisco Secure Email Gateway. An unauthenticated remote attacker can write files to the operating system and run commands as root by sending an email with a crafted attachment, per SecurityWeek and The Hacker News
Who is exposed: Physical and virtual gateways where File Reputation is enabled in a mail policy and Content Scanner Tools is earlier than 25.3.0.5525. Cisco says Secure Email and Web Manager and Secure Web Appliance are not affected
Exploitation: Cisco says it learned of active attacks in October. It has contacted managed Secure Email Cloud customers whose devices showed indicators of possible compromise
Fixes: AsyncOS 15.5.5-014 and 16.5.0-780, both released September 14. Release 16.0 has no fixed build and needs migration. The Hacker News reports Content Scanner Tools 25.3.0.5525 or later installs without a software upgrade or restart. Cisco reports no workaround
Federal deadline: SecurityWeek reports CISA added the flaw to its Known Exploited Vulnerabilities catalog with an October 17 remediation date for federal agencies
Context: The Hacker News notes a September SQL injection flaw in the same product, CVE-2026-76461, was also exploited and added to the KEV catalog. Cisco has not linked the two
Not confirmed: How many devices are compromised, who is behind the attacks, and whether any attacker obtained root. Cisco's advisory does not say

What happened

Cisco published the advisory this week and confirmed attacks against the email gateway, a device that sits in the path of inbound mail. The fixed software had been available since September 14, so the exploitation window overlaps a period when a fix existed.

The Hacker News reports Cisco's own check is a log search for "System Security" entries on each cluster member, and that a clean result does not rule out compromise. Cisco's guidance for confirmed compromise on virtual appliances is to preserve forensic data, deploy a new VM on a fixed release, rebuild configuration without importing old files and renew credentials and cryptographic material.

Evidence

Verified against independently fetched sources:
1. The Hacker News, Cisco warns of active attacks on a critical Secure Email Gateway flaw
2. SecurityWeek, Cisco warns of exploited Secure Email Gateway vulnerability

What this means for your team

An email gateway sees every inbound message and usually holds directory and mail-flow credentials. Root on that box is a position from which to read mail, pivot to identity systems and tamper with the logs that would show it.

The timing is the planning lesson. A fix existed for about a month before the advisory described attacks. Treat the date a patch ships as the start of your exposure clock for internet-facing appliances, not the date the vendor publishes an advisory.

Two exploited flaws in the same product within a month is a pattern worth a line in your risk register, even though Cisco has not tied them together.

Action checklist

  1. Inventory every Secure Email Gateway, including virtual and cloud-managed units, and record both the AsyncOS release and the Content Scanner Tools version. Check whether File Reputation is enabled in any mail policy.
  2. Install Content Scanner Tools 25.3.0.5525 or later now. It is the fastest fix and, per The Hacker News, needs no restart. Migrate any release 16.0 gateway to a fixed build.
  3. Run Cisco's log check on every cluster member, then review firewall and network logs for unexpected traffic to and from external addresses, because root access may have altered local logs.
  4. If compromise is suspected, preserve forensic data and rebuild virtual appliances on a fixed release. Send physical devices to Cisco TAC, and renew credentials the gateway held.

Palo Alto Networks fixes PAN-OS management interface zero-day CVE-2026-0322

HIGH · FIREWALL / MANAGEMENT PLANE
Key facts
Flaw: CVE-2026-0322, an authenticated remote code execution flaw in the PAN-OS management interface that gives root on the firewall. CVSS 4.0 base score 8.6 in the vendor advisory, which lists the attacker as needing high privileges
Affected: PA-Series and VM-Series firewalls and Panorama (virtual and M-Series). Cloud NGFW and Prisma Access are not affected
Exploitation: The advisory says very limited exploitation has been observed against the management interface and lists three IP addresses tied to related activity. The Hacker News describes them as U.S.-geolocated. Palo Alto Networks did not name the actors or the start date
Fixes: Fixed releases span PAN-OS 10.2 through 12.2, including 12.2.4, 12.1.4-h12, 11.2.4-h22 and 10.2.7-h38, among others. The advisory lists 11.2.16 with an estimated release of October 15, so confirm availability on that branch
Mitigation: No known workarounds. The vendor recommends restricting management interface access to trusted internal addresses. Threat IDs 510046 and 510047 give limited coverage and require SSL decryption
Federal deadline: The Hacker News and SecurityWeek report CISA added the flaw to the KEV catalog with an October 17 deadline, the same date as the Cisco flaw above
Dates: The vendor advisory is dated October 14. The two news outlets are dated October 15
Not confirmed: Attacker identity, when exploitation began and how many firewalls were affected

What happened

Palo Alto Networks published the advisory for a flaw it says it found internally and has seen used in limited attacks. The vulnerable component is the management interface, not the traffic-handling data plane.

The vendor's required privilege level means the flaw matters most where management access or administrator credentials are reachable. The advisory does not describe how the observed attacks obtained access.

Evidence

Verified against independently fetched sources:
1. Palo Alto Networks, CVE-2026-0322 security advisory
2. SecurityWeek, Palo Alto Networks patches PAN-OS zero-day exploited for remote code execution
3. The Hacker News, Palo Alto Networks patches PAN-OS authenticated RCE flaw amid active exploitation

What this means for your team

With no workaround, the one control you own before patching is who can reach the management interface. If it is reachable from the internet or from broad internal ranges, the patch is only part of the fix.

CISA listed this flaw and the Cisco flaw with the same federal deadline. If your remediation tiers follow KEV dates, both belong in the same sprint this week, owned by whoever runs perimeter infrastructure.

Action checklist

  1. List every PAN-OS firewall and Panorama instance, then map each version to the advisory's fixed builds and confirm the build is released before scheduling.
  2. Confirm the management interface is reachable only from a small set of trusted addresses, ideally through a single jump host.
  3. Review management logins, administrator account changes and configuration commits, and search for the three addresses in the advisory.
  4. Where SSL decryption is in place, enable Threat IDs 510046 and 510047, and treat them as partial coverage.

Microsoft and Health-ISAC disrupt phishing kits tied to more than one billion messages

NOTABLE · PHISHING-AS-A-SERVICE / CLOUD IDENTITY
Key facts
What happened: Microsoft and Health-ISAC led a court-authorized action against three phishing kits, CoGUI, Haozi and Iron Man System, and YAM Store, a Chinese-language marketplace that sold kits, compromised cloud access and infrastructure. The U.S. District Court for the Southern District of Florida authorized it, per SecurityWeek
Scale: Microsoft says the services sent more than one billion phishing emails and texts since January 2026 to more than 75,000 organizations in over 150 countries, and that the operators made tens of millions of dollars. About 17,000 unique domains were used, per The Hacker News
Actions taken: Domains seized and redirected, more than 20,000 accounts tied to YAM Store disabled, and more than 100,000 domains blocked, per SecurityWeek
Method: The Hacker News reports operators took over compromised or fraudulently obtained Microsoft and Google cloud tenants, abused Microsoft Graph permissions and SharePoint and OneDrive storage to host phishing, used fake CAPTCHA gates and captured credentials in real time
Attribution: Microsoft tracks YAM Store as Storm-3119 and the kit developers as Storm-2990, and assesses with high confidence the developers are Chinese-speaking
Partners: Japan Cybercrime Control Center, Resecurity, Cloudflare, Oracle, Proofpoint and TRM Labs, plus law enforcement, per the outlets reviewed
Not confirmed: Arrests, and whether the operators will rebuild. The outlets reviewed do not report either. Template counts (about 2,300) and country counts (130 to 139) vary by source

What happened

Microsoft announced the action on October 14. The investigation began in late 2025 after Japan's National Police Agency raised campaigns targeting people in Japan. Health-ISAC joined the civil case because the infrastructure was repeatedly used against hospitals and healthcare organizations.

The notable detail is the supply chain. The marketplace sold access to real cloud tenants and storage, so many of the phishing links came from legitimate-looking infrastructure.

Evidence

Verified against independently fetched sources:
1. SecurityWeek, Microsoft disrupts Chinese-language phishing market behind 1B+ messages
2. The Hacker News, Microsoft and partners disrupt three phishing kits tied to over one billion scam messages

What this means for your team

A takedown raises the cost for the operators but does not remove the demand. When stolen tenants are a product, your cloud tenant can be either a target or a free hosting platform for someone else's phishing.

Filtering on sender reputation is weaker against mail and links that originate from real Microsoft or Google tenants. Identity controls and tenant monitoring carry more of the load.

Action checklist

  1. Review third-party application consents and Microsoft Graph permissions in your tenant, and remove any you cannot tie to a business owner.
  2. Check for unusual SharePoint and OneDrive sharing and for mass outbound link activity from your own accounts.
  3. Require phishing-resistant authentication for administrators and anyone with tenant-level roles.
  4. If you are in healthcare, share what you find through Health-ISAC. Peers in other sectors can compare notes in the CISO Platform community.

Attackers backdoor Nextcloud servers through the Collabora CODE app, CVE-2026-110447

HIGH · COLLABORATION / SELF-HOSTED
Key facts
Flaw: CVE-2026-110447, an OS command injection flaw in the richdocumentscode app (Collabora Online, Built-in CODE Server). CVSS 4.0 score 9.2, per SecNews.gr. Unauthenticated requests can run commands while the app's embedded document server is not running, such as after a restart or upgrade
Affected: richdocumentscode versions before 26.4.303. Standalone Collabora Online servers are not affected
Exploitation: Attackers planted PHP web shells (files named shx*.php), tried to send system details to a webhook.site address, and on one server searched for password manager and cryptocurrency wallet files. Three administrators reported compromises. No root access is reported
Timeline: First logged attack September 10. Fix shipped in 26.4.303 on September 11. Advisory and CVE published October 14, per The Hacker News
Fix: Update to 26.4.303 or later, or remove the app. The Hacker News reports Collabora says disabling the app does not close the flaw. Some administrators report the later 26.4.402 release fails to start or open documents on their setups
Version trap: The Hacker News reports Nextcloud 29 and older cannot run the fixed version and the app store still serves them an affected build
Not confirmed: How many servers are compromised and who is behind the attacks. No group is attributed

What happened

An administrator who runs several Nextcloud servers filed the first public report, and two others described similar compromises. The attacks reportedly ran with web service privileges, which on a file collaboration server still reaches the documents users store there.

The fix shipped a day after the first logged attack, but the formal advisory and CVE came about a month later. Administrators who did not update in that gap had little reason to treat it as urgent.

Evidence

Verified against independently fetched sources:
1. The Hacker News, Attackers backdoor Nextcloud servers through a critical flaw in Collabora's CODE app
2. SecNews.gr, CVE-2026-110447: Collabora CODE flaw under active exploitation

What this means for your team

Self-hosted collaboration platforms hold contracts, board material and credentials, and they often sit outside the standard patch pipeline. A silent fix followed by a late advisory is exactly the case where inventory is the control that works.

Embedded components inherit the trust of the host application. A bundled document server is an attack surface even if no one on your team chose to deploy it.

Action checklist

  1. Find every Nextcloud instance, including ones run by business units, and check whether richdocumentscode is installed and at what version.
  2. Update to 26.4.303 or later or remove the app. Disabling it is not enough. If you run Nextcloud 29 or older, plan an upgrade rather than trusting the app store build.
  3. Search web roots for shx*.php files and review access logs back to September 10. Treat any match as a breach and start incident response.
  4. Review what the web service account can read, and rotate any secrets stored on the server.

Also notable

  • DPRK IT worker laptop farms: The Department of Justice charged 14 people, and two U.S. nationals were arrested. Raids reportedly shut down more than 40 laptop farms in Texas and seized about 187 computers. The scheme ran from about 2019 to July 2025, reached more than 70 U.S. employers and generated roughly $3 million, per the outlets. The control gap is remote-hire identity verification. Sources: SecurityWeek and The Hacker News.
  • curl: CVE-2026-92392 is a high-severity use-after-free in HTTP/2 upload handling in curl and libcurl 8.12.0 through 8.22.0. Fixed in 8.23.0, released October 14, with backports to 8.14.3, 8.16.2 and 8.20.2. No attacks are reported and the researcher's code execution demonstration required strict conditions. Embedded copies in third-party software are the inventory problem. Source: The Hacker News (single source reviewed).
  • Booz Allen AI test in an OT lab: In an isolated lab, AI agents given no prior knowledge reportedly met all objectives in eight scenarios and moved from a perimeter compromise to actions inside an industrial control network in just over 16 minutes. The lab may not match enterprise OT. Source: SecurityWeek (single source reviewed).
  • Ivanti Connect Secure follow-up: CISA added CVE-2026-104935 to the KEV catalog with an October 16 federal deadline, per SC Media. Honeypot exploitation counts differ by outlet (105 in one headline, 108 in the SC Media report). See yesterday's briefing for the full chain. Source: SC Media.
  • H&M: ShinyHunters claims it accessed H&M internal databases covering inventory, logistics and sales analytics. H&M told The Times it is aware of the claims, is investigating and has found no sign so far that its data environment was compromised. The claim is unverified. Reported October 13. Source: Cybernews (single source reviewed).

Frequently asked questions

What is CVE-2026-76519?

A critical flaw (CVSS 9.8) in Cisco Secure Email Gateway that lets an unauthenticated attacker write files and run commands as root through a crafted email attachment. Cisco confirmed active attacks. Gateways with File Reputation enabled and Content Scanner Tools earlier than 25.3.0.5525 are exposed.

How do I fix the Cisco Secure Email Gateway flaw?

Install Content Scanner Tools 25.3.0.5525 or later, or move to AsyncOS 15.5.5-014 or 16.5.0-780. Release 16.0 has no fixed build and requires migration. Cisco reports no workaround, and a clean log check does not rule out compromise.

What is CVE-2026-0322 in PAN-OS?

An authenticated remote code execution flaw in the PAN-OS management interface that gives root on the firewall. Palo Alto Networks reports very limited exploitation. It affects PA-Series, VM-Series and Panorama. The vendor recommends restricting management access to trusted internal addresses and upgrading.

What did Microsoft and Health-ISAC disrupt?

Three phishing kits, CoGUI, Haozi and Iron Man System, and the YAM Store marketplace. Microsoft says they sent more than one billion phishing messages since January 2026 to more than 75,000 organizations in over 150 countries.

How do I check for the Nextcloud Collabora CODE compromise?

Confirm the richdocumentscode app version, update to 26.4.303 or later or remove it, search for shx*.php files in web directories and review access logs back to September 10, 2026.

What should a CISO do first this week?

Patch the Cisco gateways and PAN-OS firewalls ahead of the October 17 federal deadline, restrict management interfaces, inventory self-hosted Nextcloud servers, and review third-party application permissions in your cloud tenant.

CISO Platform Breach Intelligence Team
Curated by Pritha Aash, Community Head, CISO Platform

Corrections and takedown requests: CISO Platform is committed to accuracy and fairness. If any detail in this briefing is inaccurate, or if you represent an affected organization and would like a correction or removal, please contact us at pritha.aash@cisoplatform.com and we will review your request promptly.

★
★
★
★
★
Votes: 0
E-mail me when people leave their comments –

You need to be a member of CISO Platform to add comments!

Join CISO Platform

Join The Community Discussion