Why Should I Care? β 2026-10-14 | π΄ 5 HIGH Β· π‘ 5 MEDIUM Β· π΅ 26 RADAR Β· βͺ 119 FILTERED
π Briefing β 2026-10-14
36 vendor intel items scanned | π΄ 5 HIGH | π‘ 5 MEDIUM | π΅ 26 RADAR | βͺ 119 FILTERED
π΄ Critical β action required:
- CISA Adds Two Known Exploited Vulnerabilities to Catalog (CVE-2026-96951, CVE-2026-104935) β Yes, if you run Microsoft Windows or Ivanti Connect Secure: these vulnerabilities are actively exploited in the wild and grant attackers full control or unauthorized access.
- CISA Urges Patching Newly Exploited Vulnerability in Ivanti Connect Secure (CVE-2026-104935) β Yes, if you run Ivanti Connect Secure 25.1.1.1 or prior, 25.1.2.2 or prior, or 22.8R2.4 or prior: this is a critical, actively exploited zero-day that allows unauthenticated remote code execution, leading to credential theft and data exfiltration.
- Multiple Arbitrary File Read/Write via SQL Injection β Yes, if you run FortiAnalyzer, FortiManager, or their Cloud variants: a low-privileged user can execute arbitrary commands and read/write files on the system.
- Improper access control and session management in the web management interface β Yes, if you run FortiSandbox: A remote unauthenticated attacker can bypass authentication and gain full administrative control of your sandbox management interface.
- Arbitrary file write in grpc phQueryMaster β Yes, if you run FortiSIEM: an unauthenticated attacker can write arbitrary files to the system, leading to full remote code execution.
Everything else can wait.
π‘ Medium β review when time permits:
- Inductive Automation Ignition Software β Yes, if you run Ignition 8.1.0-8.1.54 or 8.3.0-8.3.8: an authenticated low-level user can escalate to Designer privileges and execute arbitrary code on the Gateway.
- arbitrary command execution in fgfmsd daemon β Yes, if you run FortiManager in a cluster configuration: a remote attacker who has compromised a cluster node's certificate and private key can execute arbitrary OS commands on the management server.
- SQL command injection in webmail β Yes, if you run FortiMail: An authenticated user can execute arbitrary SQL commands, potentially leading to data theft or system compromise.
- GUI Denial of Service β Yes, if you run affected FortiOS versions: an unauthenticated attacker can crash your HTTPS management interface, locking you out of your firewall's admin panel.
- Fabric API Authentication bypass β Yes, if you run FortiSandbox or FortiSandbox Cloud: an unauthenticated attacker can bypass authentication on specific API endpoints using a hard-coded key, potentially exposing sandbox analysis data or configuration.
π΅ 15 items on the radar β see below β
Why Should I Care? π΄ HIGH β Handle Now
CISA Adds Two Known Exploited Vulnerabilities to Catalog
CISA Advisories [CISA KEV] | CVE-2026-96951, CVE-2026-104935
β Why Should I Care?
Yes, if you run Microsoft Windows or Ivanti Connect Secure: these vulnerabilities are actively exploited in the wild and grant attackers full control or unauthorized access.
π― Affected versions: Microsoft Windows (unspecified versions), Ivanti Connect Secure (unspecified versions)
π In plain English:
These are critical flaws that allow hackers to execute arbitrary commands on your Windows systems or bypass login security on Ivanti VPNs. An attacker could use these to steal data, install malware, or move laterally through your network without being detected.
π§ Prerequisites:
- Running an unpatched version of Microsoft Windows
- Running an unpatched version of Ivanti Connect Secure
- Exposure to network traffic or user interaction that triggers the vulnerability
β± Urgency: Immediate patching is required because CISA has confirmed active exploitation, meaning attackers are currently using these flaws to compromise systems.
π‘ Context: The root cause involves improper input validation in Windows command handling and flawed authentication logic in Ivanti Connect Secure that allows spoofing.
CISA Urges Patching Newly Exploited Vulnerability in Ivanti Connect Secure
CISA Advisories [CISA KEV] | CVE-2026-104935
β Why Should I Care?
Yes, if you run Ivanti Connect Secure 25.1.1.1 or prior, 25.1.2.2 or prior, or 22.8R2.4 or prior: this is a critical, actively exploited zero-day that allows unauthenticated remote code execution, leading to credential theft and data exfiltration.
π― Affected versions: 25.1.1.1 and prior, 25.1.2.2 and prior, 22.8R2.4 and prior
Not affected: Versions newer than 25.1.1.1, 25.1.2.2, and 22.8R2.4 (specifically the latest patched versions referenced in Ivanti's advisory)
π In plain English:
This is a backdoor that lets hackers bypass login entirely and run any command they want on your network gateway. An attacker could connect from anywhere, steal your admin passwords, and silently copy sensitive files without leaving a trace in standard logs.
π§ Prerequisites:
- Running Ivanti Connect Secure version 25.1.1.1 or prior, 25.1.2.2 or prior, or 22.8R2.4 or prior
- The appliance is exposed to the internet or an untrusted network
- No specific user interaction or valid credentials are required
β± Urgency: Immediate patching is required because this vulnerability is being actively exploited in the wild for credential theft and data exfiltration.
β Fixed in: Latest version as outlined in Ivanti Security Advisory for CVE-2026-104935
π‘ Context: The root cause is an authentication bypass flaw that allows remote, unauthenticated execution of arbitrary commands, with malicious activity executing in memory to evade standard detection tools.
Multiple Arbitrary File Read/Write via SQL Injection
Fortinet PSIRT | CVSS 7.9
β Why Should I Care?
Yes, if you run FortiAnalyzer, FortiManager, or their Cloud variants: a low-privileged user can execute arbitrary commands and read/write files on the system.
π― Affected versions: FortiAnalyzer, FortiManager, FortiManager Cloud, and FortiAnalyzer Cloud (specific versions not listed in provided text, but implied to be pre-patch releases)
π In plain English:
The system has a flaw where it doesn't properly sanitize user input before sending it to the database. An attacker with a basic account can trick the system into running malicious database commands, allowing them to steal data, modify files, or take full control of the server. For example, a low-level admin could inject code to read sensitive configuration files or create a backdoor account.
π§ Prerequisites:
- Attacker must have a low-privileged user account on the affected Fortinet product
- Attacker must be able to send crafted HTTP requests to the application
β± Urgency: Patch immediately because this allows low-privileged users to escalate to full system compromise via SQL injection.
π‘ Context: The root cause is improper neutralization of special elements in SQL commands (CWE-89), allowing injection of malicious SQL syntax.
Improper access control and session management in the web management interface
Fortinet PSIRT | CVSS 9.6
β Why Should I Care?
Yes, if you run FortiSandbox: A remote unauthenticated attacker can bypass authentication and gain full administrative control of your sandbox management interface.
π― Affected versions: FortiSandbox versions prior to the patched release (specific version ranges not listed in provided text, assume all current versions until patched)
π In plain English:
The login screen for your FortiSandbox is broken. An attacker doesn't need a password or a valid account; they can send a specific web request that tricks the system into thinking they are already logged in as an administrator. This allows them to change settings, view logs, or potentially deploy malicious configurations to your network.
π§ Prerequisites:
- Network access to the FortiSandbox web management interface (GUI)
- No valid credentials required
β± Urgency: Critical: Patch immediately as this allows full remote compromise of the security appliance without any authentication.
π‘ Context: The root cause is improper access control and session management logic that allows an attacker to forge a valid session state via crafted HTTP requests.
Arbitrary file write in grpc phQueryMaster
Fortinet PSIRT | CVSS 9.4
β Why Should I Care?
Yes, if you run FortiSIEM: an unauthenticated attacker can write arbitrary files to the system, leading to full remote code execution.
π― Affected versions: FortiSIEM versions prior to the patched release (specific version ranges not listed in provided text, typically all versions prior to the fix)
π In plain English:
The software allows anyone on the network to save files anywhere on the server without logging in. An attacker could use this to drop a malicious script or binary and then execute it, taking full control of your security monitoring system.
π§ Prerequisites:
- Network access to the FortiSIEM gRPC service
- No authentication required
β± Urgency: Critical: Patch immediately as this allows unauthenticated remote code execution on a critical security infrastructure component.
π‘ Context: The gRPC phQueryMaster service fails to properly validate file paths, allowing path traversal to write files outside the intended directory.
Why Should I Care? π‘ MEDIUM (5)
Inductive Automation Ignition Software
CISA Advisories | CVSS 8.8 | CVE-2026-92817
β Why Should I Care?
Yes, if you run Ignition 8.1.0-8.1.54 or 8.3.0-8.3.8: an authenticated low-level user can escalate to Designer privileges and execute arbitrary code on the Gateway.
π― Affected versions: 8.1.0-8.1.54, 8.3.0-8.3.8
Not affected: 8.1.55+, 8.3.9+
π In plain English:
The system fails to properly check if a user has the right level of access when they switch from a basic view to a design mode. An attacker with a basic login could trick the system into thinking they are an administrator, allowing them to run malicious code on the server.
π§ Prerequisites:
- Authenticated low-level user account
- Access to a Vision project
- Valid Vision session identifier
β± Urgency: High urgency because it allows remote code execution on critical infrastructure gateways with only low-level credentials.
β Fixed in: 8.1.55, 8.3.9
π‘ Context: The Perspective module incorrectly validates session privileges, allowing a Vision session ID to be used to gain Designer-level access.
arbitrary command execution in fgfmsd daemon
Fortinet PSIRT | CVSS 5.3
β Why Should I Care?
Yes, if you run FortiManager in a cluster configuration: a remote attacker who has compromised a cluster node's certificate and private key can execute arbitrary OS commands on the management server.
π― Affected versions: FortiManager versions prior to the patched releases listed in the advisory (specific version ranges not explicitly detailed in the provided text, but generally applies to clustered deployments).
Not affected: FortiManager instances running in standalone mode (non-clustered) are not affected by this specific cluster-impersonation vector.
π In plain English:
This is a backdoor-like vulnerability where the system trusts any device that presents a valid cluster node identity. An attacker who steals the digital ID (certificate and key) of one node in your FortiManager cluster can pretend to be that node and send malicious instructions to the main server, allowing them to run any command they want on the management system.
π§ Prerequisites:
- FortiManager must be configured in a cluster mode.
- Attacker must possess the valid certificate and private key of a legitimate FortiManager cluster node.
- Attacker must be able to send network packets to the fgfmd daemon.
β± Urgency: Medium urgency; while the impact is high (RCE), the attack requires prior compromise of cluster node credentials, making it a secondary exploitation step rather than a direct remote exploit.
β Fixed in: Refer to Fortinet PSIRT FG-IR-26-177 for specific fixed version numbers (typically the latest GA release for the affected major version).
π‘ Context: The root cause is improper input validation in the fgfmd daemon, which fails to sufficiently verify the origin of commands from cluster peers beyond basic certificate presentation, allowing OS command injection.
SQL command injection in webmail
Fortinet PSIRT | CVSS 7.7
β Why Should I Care?
Yes, if you run FortiMail: An authenticated user can execute arbitrary SQL commands, potentially leading to data theft or system compromise.
π― Affected versions: FortiMail versions prior to the patched release (specific version ranges not listed in provided text)
π In plain English:
The webmail interface has a flaw where user input isn't properly sanitized, allowing an attacker to sneak in malicious database commands. An attacker with valid login credentials could craft a special request to read, modify, or delete data in the mail server's database, or potentially gain control of the system.
π§ Prerequisites:
- Attacker must have valid authenticated access to the FortiMail webmail interface
- Attacker must be able to send specifically crafted HTTP requests
β± Urgency: High urgency because it allows authenticated attackers to execute unauthorized code or commands, posing a significant risk to data integrity and system security.
π‘ Context: The root cause is improper neutralization of special elements used in SQL commands (SQL Injection) in the FortiMail webmail component.
GUI Denial of Service
Fortinet PSIRT | CVSS 6.7
β Why Should I Care?
Yes, if you run affected FortiOS versions: an unauthenticated attacker can crash your HTTPS management interface, locking you out of your firewall's admin panel.
π― Affected versions: FortiOS 7.6.0, 7.4.0, 7.2.0, 7.0.0, 6.4.0, 6.2.0, 6.0.0 (specific minor versions listed in advisory)
Not affected: FortiOS 7.6.1, 7.4.1, 7.2.1, 7.0.1, 6.4.1, 6.2.1, 6.0.1 and later
π In plain English:
This is a bug where the firewall's web management page can be crashed by sending it a weirdly formatted request. It's like someone throwing a rock through the window of your control room; they can't steal the keys, but they can break the lights and make it impossible for you to enter the room to fix things. An attacker could send a specific HTTP request to your firewall's IP address, causing the web interface to stop responding until you reboot the device.
π§ Prerequisites:
- Network access to the FortiGate device's management interface (HTTPS)
- No authentication required
- Ability to send crafted HTTP requests
β± Urgency: Patch immediately if your management interface is exposed to the internet or untrusted networks, as this allows remote attackers to disrupt your ability to manage the device.
β Fixed in: 7.6.1, 7.4.1, 7.2.1, 7.0.1, 6.4.1, 6.2.1, 6.0.1
π‘ Context: The root cause is an improper locking mechanism (CWE-667) in the HTTP server component that fails to handle concurrent or malformed requests safely, leading to a crash.
Fabric API Authentication bypass
Fortinet PSIRT | CVSS 4.9
β Why Should I Care?
Yes, if you run FortiSandbox or FortiSandbox Cloud: an unauthenticated attacker can bypass authentication on specific API endpoints using a hard-coded key, potentially exposing sandbox analysis data or configuration.
π― Affected versions: FortiSandbox and FortiSandbox Cloud versions prior to the patched release (specific version numbers not listed in provided text, but implied to be current versions before the fix)
π In plain English:
The system has a secret key written into the code that everyone can guess or find, allowing anyone to log in without a password. An attacker could send a special web request to view files you analyzed in the sandbox or change settings, effectively bypassing your security controls.
π§ Prerequisites:
- Running FortiSandbox or FortiSandbox Cloud
- Attacker has network access to the API endpoints
- No additional authentication required due to hard-coded key
β± Urgency: Medium urgency because it requires network access to the API but allows unauthenticated access to sensitive data, posing a risk of information disclosure.
π‘ Context: The root cause is the use of a hard-coded cryptographic key in the API authentication mechanism, which fails to provide unique or secure credentials for each instance.
Why Should I Care? π΅ On the Radar (26)
- SAP Discloses Critical CVSS 10 Flaw in Cloud SDK for Python Allowing OAuth2 Credential Theft (The Hacker News) β A critical flaw (CVE-2026-89111) in SAP's Python SDK allows attackers to manipulate the tenant subdomain to redirect your application's OAuth2 client ID and secret to an attacker-controlled server. This effectively steals your credentials, granting the attacker access to your SAP BTP services. For example, if your Python app on BTP uses the SDK to authenticate, an attacker who can influence the tenant name can intercept your secret.
- Attacker Chains Three Ivanti Connect Secure Zero-Days for Root Access, Only Two Patched So Far (The Hacker News) β Attackers are actively using a combination of three bugs to take full control of Ivanti VPN appliances without needing a password. Even if you have applied the latest available patches, you are still vulnerable to one of the bugs (privilege escalation) because Ivanti has not released a fix for it yet. This means your network perimeter is currently compromised or at imminent risk of being compromised.
- Microsoft: Windows 11 24H2 Home and Pro reached end of support (BleepingComputer) β Windows 11 24H2 Home and Pro no longer receive security updates. Any device still on this version is vulnerable to new exploits. For example, if a zero-day is released next month, your 24H2 workstations will remain exploitable indefinitely.
- GitLab Patches CVSS 10.0 Flaw That Could Let Unauthenticated Users Run Server Code (The Hacker News) β GitLab has a critical flaw (CVE-2026-97229) that allows anyone on the internet to run arbitrary code on your server without logging in. This happens by sending a specially crafted regular expression in a CI/CD configuration. If exploited, an attacker could steal secrets, modify code, or take full control of your infrastructure.
- Attackers Heat French Supermarket Fridges Up to 50Β°C by Hijacking Danfoss Controllers (The Hacker News) β Attackers remotely hijacked refrigeration controllers in hundreds of French supermarkets, heating fridges to 50Β°C and forcing the disposal of millions of euros in food. This demonstrates that legacy ICS devices with internet-exposed management interfaces are a critical attack vector for physical disruption and financial damage, not just data theft.
- Windows 11 KB5127648 & KB5126653 cumulative updates released (BleepingComputer) β Microsoft has released the October 2026 security updates for Windows 11 (KB5127648 and KB5126653). These updates are required to patch security holes that could be exploited by attackers. While they also include minor UI improvements like new emojis and File Explorer tweaks, the primary value is closing security gaps. Failing to install them leaves your systems exposed to known vulnerabilities.
- Microsoft releases Windows 10 KB5126657 extended security update (BleepingComputer) β This update delivers the critical security fixes from the October 2026 Patch Tuesday, which addressed a record-breaking 1,411 vulnerabilities, including one actively exploited zero-day. If you are still on Windows 10, you are likely in a legacy support scenario (LTSC/ESU), and failing to apply this leaves you exposed to known, potentially exploited flaws. Additionally, it includes a known issue that can break USB audio devices, which may impact call centers or specialized hardware setups.
- OpenAI Employee Arrested for Allegedly Planting Hidden Recorder in Executive Conference Room (The Hacker News) β This is a news report on a criminal arrest involving physical security and espionage, not a software vulnerability or infrastructure failure.
- Anthropic warns self-replicating AI could spread to your smart devices (BleepingComputer) β Anthropic's warning highlights a theoretical future risk of AI-driven botnets on IoT devices, but no active exploitation or specific CVE is currently reported.
- Ghostscript flaw lets hackers hijack servers via PDF uploads (BleepingComputer) β News article reporting on a known CVE (CVE-2026-39919) and vendor patch, falling under the INFO rule for news items.
- OpenAI engineer arrested for allegedly planting spy device in office (BleepingComputer) β News report on a corporate security incident involving personnel misconduct, not a technical vulnerability.
- Part 2 β Cybersecurity journeys: What even is this job? (AWS Security Blog) β Career guidance article with no security impact or technical advisory content.
- The Noise Inside the Network: How an Unsophisticated Attack Exposed a Third-Party Risk (Palo Alto Unit 42) β Blog post analyzing a specific incident and third-party risk, not a new vulnerability disclosure.
- GitHub Tightens npm Publishing After Flaw Left 1,500+ Packages Vulnerable to Hijacking (The Hacker News) β Vendor mitigation for a configuration flaw with no evidence of active exploitation.
- Docker Fixes Critical Two-Year-Old Hub Flaw That Let One User Push Images as Another (The Hacker News) β News report on a vendor-patched flaw with no evidence of exploitation.
βͺ 119 low-priority items filtered.
π¦ Aggregated and triaged by Donna AI | Sources: 8 vendor feeds | CISA KEV