Why Should I Care? β 2026-11-11 | π΄ 19 HIGH Β· π‘ 75 MEDIUM Β· π΅ 15 RADAR Β· βͺ 10 FILTERED
π Briefing β 2026-11-11
119 vendor intel items scanned | π΄ 19 HIGH | π‘ 75 MEDIUM | π΅ 15 RADAR | βͺ 10 FILTERED
π΄ Critical β action required:
- Incorrect authorization (CVE-2026-126155) β Yes, if you run FortiMail 7.2, 7.4, 7.6, or 8.0: An unauthenticated attacker can hijack any IBE user's session to read or send their encrypted emails.
- Incorrect authorization (CVE-2026-126251) β Yes, if you run FortiAuthenticator 8.0.0-8.0.3 or 6.6.0-6.6.10: an unauthenticated attacker can bypass authentication and gain unauthorized access to your identity infrastructure.
- Improper neutralization of special elements used in an os command (CVE-2026-126148) β Yes, if you run FortiMail 8.0.0 through 8.0.1 or 7.6.0 through 7.6.6: an unauthenticated attacker can execute arbitrary commands on your mail server via crafted HA join-request packets.
- Improper neutralization of special elements used in an SQL command (CVE-2026-126164) β Yes, if you run FortiMail 8.0, 7.6, or 7.4: an unauthenticated attacker can execute arbitrary code or commands via SQL injection, leading to full system compromise or denial of service.
- Improper limitation of a pathname to a restricted directory (CVE-2026-126147) β Yes, if you run FortiMail 7.2, 7.4, 7.6, or 8.0: an unauthenticated attacker can execute arbitrary code on your mail server via a path traversal vulnerability in the API.
- FGFM elevation before registration password is verified β CVSS 9.0 β critical severity, immediate attention required.
- Authorization bypass through user-controlled key in dmserver β CVSS 9.1 β critical severity, immediate attention required.
- Authorization bypass through user-controlled key β CVSS 9.1 β critical severity, immediate attention required.
- Authentication bypass by primary weakness β CVSS 9.3 β critical severity, immediate attention required.
- Authentication bypass β CVSS 9.0 β critical severity, immediate attention required.
- CVE-2026-85839 Microsoft WDAC OLE DB provider for SQL Remote Code Execution Vulnerability (CVE-2026-85839) β Yes, if you run Windows 10/11 or Server 2016+ with the WDAC OLE DB provider for SQL enabled: a remote attacker can execute arbitrary code on your system via a crafted SQL connection.
- CVE-2026-85838 Microsoft WDAC OLE DB provider for SQL Remote Code Execution Vulnerability (CVE-2026-85838) β Yes, if you run any Windows version with the WDAC OLE DB provider for SQL enabled: a remote attacker can execute arbitrary code on your system without authentication.
- CVE-2026-94511 SQL Server Remote Code Execution Vulnerability (CVE-2026-94511) β Yes, if you run any supported version of Microsoft SQL Server: an authenticated user with write permissions can execute arbitrary code on the database server.
- CVE-2026-114169 Windows Raw Image Extension Remote Code Execution Vulnerability (CVE-2026-114169) β Yes, if you run Windows 10 or 11: A remote attacker can execute arbitrary code on your system by tricking you into opening a malicious raw image file.
- CVE-2026-114164 HEVC Video Extensions Remote Code Execution Vulnerability (CVE-2026-114164) β Yes, if you run Windows 10 or 11: A remote attacker can execute arbitrary code on your system by tricking you into opening a malicious HEVC video file.
- CVE-2026-66801 Microsoft Xbox Remote Code Execution Vulnerability β Unauthenticated remote code execution vulnerability in a widely deployed consumer and enterprise device.
- CVE-2026-109467 Azure DevOps Server Remote Code Execution Vulnerability β Remote code execution vulnerability in Azure DevOps Server allows attackers to compromise CI/CD pipelines.
- CVE-2026-109463 Windows DNS Remote Code Execution Vulnerability β Unauthenticated remote code execution in Windows DNS is a critical risk for domain controllers and network infrastructure.
- CVE-2026-108411 Microsoft ODBC Driver Remote Code Execution Vulnerability β Remote Code Execution vulnerability in a network-facing component (ODBC Driver) allows unauthorized attackers to execute code over a network.
Everything else can wait.
π‘ Medium β review when time permits:
- Cisco Identity Services Engine Authorization Bypass Vulnerabilities β Yes, if you run Cisco ISE or ISE-PIC versions 3.2 through 3.5: An authenticated administrator can bypass permission checks to modify file descriptions, potentially altering audit logs or system metadata.
- Incorrect authorization β Vendor advisory for a high-severity (CVSS 8.7) privilege escalation vulnerability in Fortinet management products.
- Improper privilege management β CVSS 7.2 privilege escalation vulnerability in Fortinet management and analytics platforms requires patching.
- Improper privilege management β CVSS 7.2 privilege escalation vulnerability in FortiAuthenticator requires immediate remediation.
- Improper neutralization of special elements used in an os command in GUI β CVSS 7.2 OS command injection vulnerability in Fortinet management and SASE products allows unauthorized code execution.
- Improper neutralization of special elements used in an os command β CVSS 7.2 OS command injection vulnerability in FortiMail allows root-level code execution by authenticated admins.
- CVE-2026-62903 Microsoft.Data.SqlClient Information Disclosure Vulnerability β Information disclosure vulnerability via authentication bypass in SQL Server client libraries.
- CVE-2026-100223 Windows NTFS Elevation of Privilege Vulnerability β Elevation of privilege vulnerability allowing sandbox escape via reparse points.
- CVE-2026-83684 Windows iSCSI Target Service Information Disclosure Vulnerability β Vendor advisory for an information disclosure vulnerability in Windows iSCSI Target Service with no reported active exploitation.
- CVE-2026-85801 Windows Local Security Authority Subsystem Service (LSASS) Elevation of Privilege Vulnerability β Vendor advisory for a local privilege escalation vulnerability in LSASS requiring an authorized attacker.
- ... and 65 more
π΅ 15 items on the radar β see below β
π΄ HIGH β Handle Now
Incorrect authorization
Fortinet PSIRT | CVSS 9.1 | CVE-2026-126155
β Why Should I Care?
Yes, if you run FortiMail 7.2, 7.4, 7.6, or 8.0: An unauthenticated attacker can hijack any IBE user's session to read or send their encrypted emails.
π― Affected versions: FortiMail 8.0.0 through 8.0.1, 7.6.0 through 7.6.6, 7.4.0 through 7.4.8, 7.2 all versions
Not affected: FortiMail 8.0.2 and above, 7.6.7 and above, 7.4.9 and above, FortiMail Cloud (remediated Oct 05, 2026)
π In plain English:
This is a critical flaw in the email server's login logic that lets an outsider impersonate any user without a password. An attacker could send a specific web request to the server, trick it into thinking they are a specific employee, and then read that employee's private, encrypted emails or send messages on their behalf.
π§ Prerequisites:
- Running an affected FortiMail version (7.2, 7.4, 7.6, or 8.0)
- Internet access to the FortiMail webmail interface
- Knowledge of a valid IBE recipient email address
β± Urgency: Immediate patching is required because this allows unauthenticated remote attackers to fully compromise the confidentiality and integrity of user email data.
β Fixed in: 8.0.2, 7.6.7, 7.4.9
π‘ Context: The root cause is an Incorrect Authorization (CWE-863) flaw in the API that fails to properly validate session ownership, allowing session hijacking via crafted HTTP requests.
Incorrect authorization
Fortinet PSIRT | CVSS 9.1 | CVE-2026-126251
β Why Should I Care?
Yes, if you run FortiAuthenticator 8.0.0-8.0.3 or 6.6.0-6.6.10: an unauthenticated attacker can bypass authentication and gain unauthorized access to your identity infrastructure.
π― Affected versions: FortiAuthenticator 8.0.0 through 8.0.3, 6.6.0 through 6.6.10
Not affected: FortiAuthenticator 8.0.4 and above, 6.6.11 and above
π In plain English:
The system fails to properly check if a user is allowed to perform an action, allowing anyone to log in as an administrator without a password. An attacker could use this to steal user credentials, reset passwords, or gain full control of your identity provider, effectively compromising all systems that rely on it for login.
π§ Prerequisites:
- Network access to the FortiAuthenticator API
- Running a vulnerable version (8.0.0-8.0.3 or 6.6.0-6.6.10)
β± Urgency: Critical and immediate action is required because this is an unauthenticated remote exploit that grants full administrative control over your identity infrastructure.
β Fixed in: 8.0.4, 6.6.11
π‘ Context: The root cause is an incorrect authorization check in the API that allows bypassing authentication controls.
Improper neutralization of special elements used in an os command
Fortinet PSIRT | CVSS 9.8 | CVE-2026-126148
β Why Should I Care?
Yes, if you run FortiMail 8.0.0 through 8.0.1 or 7.6.0 through 7.6.6: an unauthenticated attacker can execute arbitrary commands on your mail server via crafted HA join-request packets.
π― Affected versions: FortiMail 8.0.0 through 8.0.1, 7.6.0 through 7.6.6
Not affected: FortiMail 7.4, 7.2
π In plain English:
The system fails to properly sanitize input from High Availability (HA) join requests, allowing an attacker to inject and run malicious system commands. For example, an attacker could send a specially crafted packet to the HA interface to gain full control of the mail server, potentially reading emails or installing malware.
π§ Prerequisites:
- Running FortiMail 8.0.0-8.0.1 or 7.6.0-7.6.6
- Attacker has network access to the FortiMail HA interface
- No authentication required
β± Urgency: Critical and immediate action is required because this is an unauthenticated remote code execution vulnerability with a CVSS score of 9.8.
β Fixed in: 8.0.2, 7.6.7
π‘ Context: The root cause is improper neutralization of special elements in OS command construction (CWE-77) within the HA join-request handling logic.
Improper neutralization of special elements used in an SQL command
Fortinet PSIRT | CVSS 9.4 | CVE-2026-126164
β Why Should I Care?
Yes, if you run FortiMail 8.0, 7.6, or 7.4: an unauthenticated attacker can execute arbitrary code or commands via SQL injection, leading to full system compromise or denial of service.
π― Affected versions: FortiMail 8.0.0 through 8.0.1, 7.6.0 through 7.6.6, 7.4.0 through 7.4.8, 7.2 all versions
Not affected: FortiMail Cloud (remediated Oct 05, 2026)
π In plain English:
The FortiMail system has a flaw where it doesn't properly sanitize user input before sending it to its database. An attacker can send specially crafted requests that trick the system into running malicious database commands. For example, an attacker could inject SQL code to delete all email records, steal user credentials, or take full control of the mail server without needing a login.
π§ Prerequisites:
- Running an affected version of FortiMail (8.0.0-8.0.1, 7.6.0-7.6.6, 7.4.0-7.4.8, or any 7.2.x)
- Attacker has network access to the FortiMail API
- No authentication required
β± Urgency: Critical and immediate action is required because this is an unauthenticated remote code execution vulnerability with a CVSS score of 9.4, allowing full system compromise.
β Fixed in: 8.0.2, 7.6.7, 7.4.9
π‘ Context: The root cause is improper neutralization of special elements used in an SQL command (CWE-89) in the FortiMail API, allowing SQL injection.
Improper limitation of a pathname to a restricted directory
Fortinet PSIRT | CVSS 9.8 | CVE-2026-126147
β Why Should I Care?
Yes, if you run FortiMail 7.2, 7.4, 7.6, or 8.0: an unauthenticated attacker can execute arbitrary code on your mail server via a path traversal vulnerability in the API.
π― Affected versions: FortiMail 8.0.0 through 8.0.1, 7.6.0 through 7.6.6, 7.4.0 through 7.4.8, 7.2 all versions
Not affected: FortiMail 8.0.2 and above, 7.6.7 and above, 7.4.9 and above
π In plain English:
The FortiMail API has a flaw where it doesn't properly check the file paths in requests, allowing an attacker to escape the intended directory and access or execute files elsewhere on the system. An attacker could send a specially crafted request to the API to run malicious commands, potentially taking full control of the mail server and intercepting or modifying email traffic.
π§ Prerequisites:
- FortiMail API is accessible from the network
- Attacker is unauthenticated (no credentials required)
- Running an affected version of FortiMail
β± Urgency: Critical and immediate action is required because this is an unauthenticated remote code execution vulnerability with a CVSS score of 9.8, allowing full system compromise without any prior access.
β Fixed in: 8.0.2, 7.6.7, 7.4.9
π‘ Context: The root cause is an improper limitation of a pathname to a restricted directory (CWE-22) in the FortiMail API, allowing path traversal to execute unauthorized code.
FGFM elevation before registration password is verified
Fortinet PSIRT
CVSS 9.0 β critical severity, immediate attention required.
π― Affected: Fortinet
Authorization bypass through user-controlled key in dmserver
Fortinet PSIRT
CVSS 9.1 β critical severity, immediate attention required.
π― Affected: Fortinet
Authorization bypass through user-controlled key
Fortinet PSIRT
CVSS 9.1 β critical severity, immediate attention required.
π― Affected: Fortinet
Authentication bypass by primary weakness
Fortinet PSIRT
CVSS 9.3 β critical severity, immediate attention required.
π― Affected: Fortinet
Authentication bypass
Fortinet PSIRT
CVSS 9.0 β critical severity, immediate attention required.
π― Affected: Fortinet
CVE-2026-85839 Microsoft WDAC OLE DB provider for SQL Remote Code Execution Vulnerability
Microsoft MSRC | CVSS 8.8 | CVE-2026-85839
β Why Should I Care?
Yes, if you run Windows 10/11 or Server 2016+ with the WDAC OLE DB provider for SQL enabled: a remote attacker can execute arbitrary code on your system via a crafted SQL connection.
π― Affected versions: Windows 10, Windows 11, Windows Server 2016, Windows Server 2019, Windows Server 2022, Windows Server 2025
π In plain English:
This is a memory corruption bug in the component that lets Windows connect to SQL databases. An attacker can send a specially formatted database request that overwrites memory, allowing them to run malicious commands on your computer. For example, an attacker could exploit this to install ransomware or steal data from a server that connects to a compromised SQL database.
π§ Prerequisites:
- Target system must have the Microsoft WDAC OLE DB provider for SQL installed
- Attacker must be able to send a crafted SQL request to the target (network access)
- User interaction may be required (e.g., opening a malicious link or file that triggers the connection)
β± Urgency: Patch immediately if you use SQL connectivity on Windows, as this allows remote code execution with high impact and low attack complexity.
β Fixed in: Windows 10 22H2 (latest cumulative update), Windows 11 23H2 (latest cumulative update), Windows Server 2016 (latest cumulative update), Windows Server 2019 (latest cumulative update), Windows Server 2022 (latest cumulative update), Windows Server 2025 (latest cumulative update)
π‘ Context: The root cause is a heap-based buffer overflow in the OLE DB provider's handling of SQL data, which fails to properly validate input size before writing to memory.
CVE-2026-85838 Microsoft WDAC OLE DB provider for SQL Remote Code Execution Vulnerability
Microsoft MSRC | CVE-2026-85838
β Why Should I Care?
Yes, if you run any Windows version with the WDAC OLE DB provider for SQL enabled: a remote attacker can execute arbitrary code on your system without authentication.
π― Affected versions: Windows 11, Windows 10, Windows Server 2022, Windows Server 2019, Windows Server 2016 (versions with WDAC OLE DB provider for SQL)
π In plain English:
There is a flaw in how Windows handles database connections that allows an attacker to send specially crafted data over the network to crash the system or run malicious programs. For example, an attacker could connect to a vulnerable SQL server and trigger this bug to install malware or steal data without needing a password.
π§ Prerequisites:
- Target system must have the Microsoft WDAC OLE DB provider for SQL installed
- Attacker must have network access to the vulnerable service
- No user interaction or authentication required
β± Urgency: Patch immediately as this allows unauthenticated remote code execution over the network.
β Fixed in: Windows 11 24H2 (latest cumulative update), Windows 10 22H2 (latest cumulative update), Windows Server 2022 (latest cumulative update), Windows Server 2019 (latest cumulative update), Windows Server 2016 (latest cumulative update)
π‘ Context: The root cause is a heap-based buffer overflow in the WDAC OLE DB provider for SQL that fails to properly validate input data from remote connections.
CVE-2026-94511 SQL Server Remote Code Execution Vulnerability
Microsoft MSRC | CVE-2026-94511
β Why Should I Care?
Yes, if you run any supported version of Microsoft SQL Server: an authenticated user with write permissions can execute arbitrary code on the database server.
π― Affected versions: All supported versions of Microsoft SQL Server (2019, 2022, 2025)
π In plain English:
A flaw in how SQL Server handles memory allows a user who already has access to the database to crash the system or run malicious commands. For example, a compromised application account or a malicious insider could use this bug to install a backdoor, steal data, or take full control of the server hosting the database.
π§ Prerequisites:
- Attacker must have valid credentials for a SQL Server user with sufficient privileges (e.g., db_owner or sysadmin).
- Network access to the SQL Server instance is required.
β± Urgency: Patch immediately if you have any exposed SQL Server instances, as this allows remote code execution by any authenticated user.
β Fixed in: SQL Server 2019 CU 13, SQL Server 2022 CU 13, SQL Server 2025 CU 1
π‘ Context: The root cause is a heap-based buffer overflow in the SQL Server engine that fails to properly validate input size before writing to memory.
CVE-2026-114169 Windows Raw Image Extension Remote Code Execution Vulnerability
Microsoft MSRC | CVSS 8.8 | CVE-2026-114169
β Why Should I Care?
Yes, if you run Windows 10 or 11: A remote attacker can execute arbitrary code on your system by tricking you into opening a malicious raw image file.
π― Affected versions: Windows 10, Windows 11 (specific versions not listed in provided text, typically all current releases prior to patch)
π In plain English:
This is a memory corruption bug in how Windows handles raw image files. An attacker could craft a malicious image file that, when opened, crashes the system or runs hidden malware, potentially giving them full control of your computer.
π§ Prerequisites:
- Attacker must deliver a malicious raw image file to the victim
- Victim must open the file with the vulnerable Windows Raw Image Extension
- No administrative privileges required for the initial attack
β± Urgency: High urgency because it allows remote code execution with high impact on confidentiality, integrity, and availability, requiring immediate patching.
π‘ Context: The root cause is a heap-based buffer overflow in the Windows Raw Image Extension component.
CVE-2026-114164 HEVC Video Extensions Remote Code Execution Vulnerability
Microsoft MSRC | CVSS 8.8 | CVE-2026-114164
β Why Should I Care?
Yes, if you run Windows 10 or 11: A remote attacker can execute arbitrary code on your system by tricking you into opening a malicious HEVC video file.
π― Affected versions: Windows 10, Windows 11 (specific versions not listed in provided text, typically all current supported releases prior to patch)
π In plain English:
The software that plays HEVC videos has a math error that lets a bad guy send a specially crafted video file. If you open it, your computer runs their malicious code instead of just playing the video. For example, an attacker could send you a 'funny cat video' in an email; when you double-click it, your PC gets infected with malware.
π§ Prerequisites:
- Attacker must deliver a malicious HEVC video file to the victim
- Victim must open or preview the file
- System must be running an unpatched version of Windows with the vulnerable Codecs Library
β± Urgency: High urgency because it allows remote code execution with no special privileges required, though user interaction is needed.
π‘ Context: The root cause is an integer underflow in the Windows Codecs Library when processing HEVC video data.
CVE-2026-66801 Microsoft Xbox Remote Code Execution Vulnerability
Microsoft MSRC
Unauthenticated remote code execution vulnerability in a widely deployed consumer and enterprise device.
π― Affected: Microsoft Xbox
Action: Apply latest Xbox system update immediately.
CVE-2026-109467 Azure DevOps Server Remote Code Execution Vulnerability
Microsoft MSRC
Remote code execution vulnerability in Azure DevOps Server allows attackers to compromise CI/CD pipelines.
π― Affected: Azure DevOps Server
Action: Apply latest Azure DevOps Server security update.
CVE-2026-109463 Windows DNS Remote Code Execution Vulnerability
Microsoft MSRC
Unauthenticated remote code execution in Windows DNS is a critical risk for domain controllers and network infrastructure.
π― Affected: Windows DNS
Action: Apply latest Windows security update to all DNS servers.
CVE-2026-108411 Microsoft ODBC Driver Remote Code Execution Vulnerability
Microsoft MSRC
Remote Code Execution vulnerability in a network-facing component (ODBC Driver) allows unauthorized attackers to execute code over a network.
π― Affected: Microsoft Windows ODBC Driver
Action: Apply Microsoft security patch immediately
π‘ MEDIUM (75)
Cisco Identity Services Engine Authorization Bypass Vulnerabilities
Cisco Security Advisories | CVSS 4.3 | CVE-2026-20183, CVE-2026-20285, CVE-2026-20286
β Why Should I Care?
Yes, if you run Cisco ISE or ISE-PIC versions 3.2 through 3.5: An authenticated administrator can bypass permission checks to modify file descriptions, potentially altering audit logs or system metadata.
π― Affected versions: Cisco ISE 3.2 and earlier, 3.3 prior to Patch 12, 3.4 prior to Patch 7, 3.5 prior to Patch 4; Cisco ISE-PIC 3.2 and earlier, 3.3 prior to Patch 12, 3.4 prior to Patch 7
Not affected: Cisco ISE 3.3 Patch 12 and later, 3.4 Patch 7 and later, 3.5 Patch 4 and later; Cisco ISE-PIC 3.3 Patch 12 and later, 3.4 Patch 7 and later
π In plain English:
The system fails to properly check if a user has the right to change certain settings, even if they are logged in as an administrator. An attacker with valid admin credentials could send a specific web request to change the descriptions of files on the system, which could be used to hide malicious activity or confuse auditors.
π§ Prerequisites:
- Attacker must have valid Administrator credentials
- Attacker must be able to send HTTP requests to the web-based management interface
β± Urgency: Patch within your standard maintenance window as this requires high-privilege access and has limited impact, but it undermines the integrity of your access control system.
β Fixed in: 3.3 Patch 12, 3.4 Patch 7, 3.5 Patch 4
π‘ Context: The root cause is the lack of server-side validation of Administrator permissions in the web-based management interface.
Incorrect authorization
Fortinet PSIRT
Vendor advisory for a high-severity (CVSS 8.7) privilege escalation vulnerability in Fortinet management products.
π― Affected: FortiManager, FortiManager Cloud, FortiAnalyzer, FortiAnalyzer Cloud, FortiAnalyzer-BigData
β Fixed in: N/A
Action: Apply the latest security patch to FortiManager and FortiAnalyzer instances.
Improper privilege management
Fortinet PSIRT
CVSS 7.2 privilege escalation vulnerability in Fortinet management and analytics platforms requires patching.
π― Affected: FortiAnalyzer, FortiAnalyzer Cloud, FortiManager, FortiManager Cloud
Action: Apply vendor patch and restrict JSON-RPC access to trusted administrators.
Improper privilege management
Fortinet PSIRT
CVSS 7.2 privilege escalation vulnerability in FortiAuthenticator requires immediate remediation.
π― Affected: FortiAuthenticator
Action: Apply vendor patch and audit privileged user accounts.
Improper neutralization of special elements used in an os command in GUI
Fortinet PSIRT
CVSS 7.2 OS command injection vulnerability in Fortinet management and SASE products allows unauthorized code execution.
π― Affected: FortiManager, FortiAnalyzer, FortiManager Cloud, FortiAnalyzer Cloud, FortiSASE-Sovereign GUI
Action: Apply vendor patch and monitor for suspicious HTTP requests.
Improper neutralization of special elements used in an os command
Fortinet PSIRT
CVSS 7.2 OS command injection vulnerability in FortiMail allows root-level code execution by authenticated admins.
π― Affected: FortiMail
Action: Apply vendor patch and restrict administrative access to trusted networks.
CVE-2026-62903 Microsoft.Data.SqlClient Information Disclosure Vulnerability
Microsoft MSRC
Information disclosure vulnerability via authentication bypass in SQL Server client libraries.
π― Affected: Microsoft.Data.SqlClient
Action: Update Microsoft.Data.SqlClient libraries to the latest patched version.
CVE-2026-100223 Windows NTFS Elevation of Privilege Vulnerability
Microsoft MSRC
Elevation of privilege vulnerability allowing sandbox escape via reparse points.
π― Affected: Windows NTFS
Action: Apply standard Windows security updates.
CVE-2026-83684 Windows iSCSI Target Service Information Disclosure Vulnerability
Microsoft MSRC
Vendor advisory for an information disclosure vulnerability in Windows iSCSI Target Service with no reported active exploitation.
π― Affected: Windows iSCSI Target Service
Action: Apply Microsoft security updates
CVE-2026-85801 Windows Local Security Authority Subsystem Service (LSASS) Elevation of Privilege Vulnerability
Microsoft MSRC
Vendor advisory for a local privilege escalation vulnerability in LSASS requiring an authorized attacker.
π― Affected: Windows Local Security Authority Subsystem Service (LSASS)
Action: Apply Microsoft security updates
CVE-2026-94513 Microsoft SQL Server Elevation of Privilege Vulnerability
Microsoft MSRC
Vendor advisory for a local privilege escalation vulnerability in SQL Server requiring an authorized attacker.
π― Affected: Microsoft SQL Server
Action: Apply Microsoft security updates
CVE-2026-0250 GlobalProtect App: Buffer Overflow Vulnerability during connection to Portal or Gateway (Severity: MEDIUM)
Palo Alto Security Advisories | CVSS 5.2 | CVE-2026-0250
β Why Should I Care?
Yes, if you run GlobalProtect App versions 6.0 through 6.3 on Windows, macOS, Android, ChromeOS, or Linux: a man-in-the-middle attacker can execute arbitrary code with SYSTEM privileges.
π― Affected versions: GlobalProtect App 6.3 < 6.3.3-h9 (Windows/macOS), < 6.3.5 (Android/ChromeOS), < 6.3.3-h15 (Linux); 6.2 < 6.2.8-h10 (Windows/macOS); 6.1 < 6.1.14 (Android/ChromeOS); 6.0 < 6.0.13 (Windows/macOS), < 6.0.15 (Android/ChromeOS/Linux); UWP App < 6.3.3-h10 (Windows)
Not affected: GlobalProtect App on iOS
π In plain English:
The GlobalProtect app has a flaw in how it handles data from the server. An attacker who can intercept the connection (like on a compromised Wi-Fi network) can send malicious data that crashes the app or runs their own code with full system control. For example, an attacker on the same network could inject a crafted packet that installs a persistent backdoor on your laptop.
π§ Prerequisites:
- Attacker must be in a position to perform a man-in-the-middle attack (e.g., same network segment or compromised upstream)
- Victim must be running an affected version of the GlobalProtect App (not iOS)
- Victim must be connected to a GlobalProtect Portal or Gateway
β± Urgency: Moderate urgency because while the impact is high (SYSTEM code execution), the attack requires a specific man-in-the-middle position and has not been seen in the wild.
β Fixed in: 6.3.3-h9, 6.3.5, 6.3.3-h15, 6.2.8-h10, 6.1.14, 6.0.13, 6.0.15, 6.3.3-h10
π‘ Context: The root cause is an out-of-bounds write (CWE-787) triggered during the processing of requests and responses exchanged between the Portal and Gateway.
Improper neutralization of special elements used in an os command
Fortinet PSIRT
CVSS 7.2 OS command injection vulnerability in FortiManager and FortiAnalyzer allows unauthorized code execution.
π― Affected: FortiManager, FortiManager Cloud, FortiAnalyzer, FortiAnalyzer Cloud
Action: Apply vendor patch and review administrative activity logs.
Improper neutralization of special elements used in an ldap query
Fortinet PSIRT
CVSS 8.2 authentication bypass in FortiAuthenticator requires specific IdP misconfiguration or compromise.
π― Affected: FortiAuthenticator
Action: Review IdP configurations and apply vendor patches.
Improper neutralization of special elements used in an SQL command in collectoragent
Fortinet PSIRT
CVSS 8.3 SQL injection in collectoragent allows authenticated attackers to execute unauthorized commands.
π― Affected: Fortinet collectoragent
Action: Apply vendor patches to collectoragent.
... and 60 more medium-priority items.
π΅ On the Radar (15)
- Attackers Exploit AhsayCBS Flaws to Deploy Sandbox-Evading Python Backdoor (The Hacker News) β Attackers are actively exploiting two critical flaws in AhsayCBS to bypass authentication and install sophisticated malware. They are not just stealing data; they are stealing the credentials stored in backup systems (like Veeam) to gain control over hypervisors and domain accounts. For example, an attacker could use stolen Veeam credentials to restore a malicious backup or lock you out of your entire infrastructure.
- CISA Warns of Active Exploitation of OpenSSH, GPUStack, FreeSWITCH, and IBM Flaws (The Hacker News) β CISA has confirmed that attackers are actively exploiting four critical flaws. The most dangerous is an OpenSSH bug (CVE-2026-123158) that allows authentication bypass on Linux systems. Attackers are already using this to gain root access, steal credentials, and install malware like the TAC Agent. If you have any of these systems exposed to the internet or internal networks, you are at immediate risk of compromise.
- Railway breach leads to $740,000 in fraudulent AI charges (BleepingComputer) β Hackers breached Railway's infrastructure, stole internal source code and customer environment variables (including API keys), and used them to rack up $740,000 in fraudulent AI charges. If your organization hosts applications on Railway or similar PaaS providers, your secrets may have been exposed and abused for financial gain or data exfiltration.
- Fireworks AI Confirms Wider Breach Affecting Customer Datasets, Model Weights and Notion AI Chats (The Hacker News) β Fireworks AI suffered a breach where attackers accessed internal credentials via a third-party cloud provider, leading to the theft of API keys, billing information, and incomplete model weights for some customers. Notion AI was also affected, with chat transcripts potentially exposed. This is not just a data leak; it's an active financial threat where stolen AI access is being resold, potentially causing you unexpected cloud bills or data exfiltration.
- Hackers Exploit Critical Xen Orchestra Auth Bypass Flaw to Gain Admin Access (The Hacker News) β Hackers are using a specific cookie trick to bypass login on Xen Orchestra, allowing them to steal admin tokens and take over the entire virtualization infrastructure. If you have this software exposed to the internet, you are likely already compromised or at imminent risk of losing control of your VMs.
- SGLang RCE Exploit Breaches NEAR AI, Steals Data From 18 Agents and About 100 Users (The Hacker News) β An attacker exploited a remote code execution flaw in SGLang to gain root-level access to NEAR AI's inference servers. Because SGLang runs inside the Trusted Execution Environment (TEE), the attacker could read confidential user requests and data that were supposed to be protected from the host OS. This resulted in the theft of user credentials, API keys, and chat histories from 18 AI agents and approximately 100 users. If you run SGLang, your inference servers are likely vulnerable to the same attack vector.
- Hackers exploit trusted AWS access to breach multiple cloud environments (BleepingComputer) β Hackers compromised a big data platform (Qubole) and used its legitimate access rights to break into dozens of other companies' AWS accounts. They stole passwords, API keys, and bulk data from S3 buckets. If you have ever granted a third-party vendor access to your AWS account via IAM roles, your credentials and data may have been exposed.
- Binance Warns of iOS Sandbox Escape That Could Expose Crypto Wallet Keys, No Fix Confirmed (The Hacker News) β Unpatched iOS sandbox escape allows cross-app data theft of sensitive credentials and keys.
- France Blames Iran for Sabotage Attack on Refrigeration Controllers at About 700 Retail Sites (The Hacker News) β Physical sabotage of critical infrastructure (refrigeration) by a state actor caused fires and significant operational loss.
- SAP Patches Four CVSS 10.0 Flaws, Including Two Enabling Remote Code Execution (The Hacker News) β CVSS 10.0 β critical severity, immediate attention required.
- Turla hackers use new Cereal malware to spy on govt Exchange servers (BleepingComputer) β Novel state-sponsored attack campaign targeting government infrastructure with no vendor advisory yet.
- N-able Takes N-central Cloud Offline After Researcher Discloses 33 Zero-Days with PoC Exploits (The Hacker News) β Active exploitation of 33 zero-days with public PoC exploits and vendor taking hosted services offline indicates critical risk.
- Fireworks AI discloses new breach after cloud provider hack (BleepingComputer) β Genuine data breach disclosure involving AI infrastructure and third-party cloud provider compromise.
- Blind Eagle Uses Fake Court Notices to Deliver Remote Access Trojan Across Latin America (The Hacker News) β Active phishing campaign targeting Latin American organizations with a known RAT lineage.
- Three China-Nexus Espionage Campaigns Linked Through Sparrow Malware Ecosystem (The Hacker News) β Report of ongoing espionage campaigns targeting critical infrastructure and government entities.
βͺ 10 low-priority items filtered.
π¦ Aggregated and triaged by Donna AI | Sources: 10 vendor feeds | CISA KEV