Why Should I Care? β 2026-10-22 | π΄ 3 HIGH Β· π‘ 0 MEDIUM Β· π΅ 12 RADAR Β· βͺ 182 FILTERED
π Briefing β 2026-10-22
15 vendor intel items scanned | π΄ 3 HIGH | π‘ 0 MEDIUM | π΅ 12 RADAR | βͺ 182 FILTERED
π΄ Critical β action required:
- Using AWS WAF to apply mitigations for PHP2Shell RCE vulnerabilities (CVE-2026-109618 and CVE-2026-109619) β CVSS 9.2 β critical severity, immediate attention required.
- Amazon Threat Intelligence identifies PHP2Shell exploitation prior to public disclosure in MadPot (CVE-2026-109618, CVE-2026-109619) β Yes, if you run any unpatched PHP version: PHP2Shell is being actively exploited in the wild, with attacks observed four days before public disclosure, meaning compromise may have already occurred.
- PHP core RCE vulnerability patches (CVE-2026-109618 and CVE-2026-109619) (CVE-2026-109618) β Yes, if you run FortiADC, FortiClient EMS, FortiDDoS, FortiSIEM, FortiMonitor, FortiClient Cloud, or FortiSOAR: these products are under investigation for a critical unauthenticated RCE vulnerability in the PHP core.
Everything else can wait.
π΅ 12 items on the radar β see below β
π΄ HIGH β Handle Now
Using AWS WAF to apply mitigations for PHP2Shell RCE vulnerabilities (CVE-2026-109618 and CVE-2026-109619)
AWS Security Blog
CVSS 9.2 β critical severity, immediate attention required.
π― Affected: AWS
Amazon Threat Intelligence identifies PHP2Shell exploitation prior to public disclosure in MadPot
AWS Security Blog | CVE-2026-109618, CVE-2026-109619
β Why Should I Care?
Yes, if you run any unpatched PHP version: PHP2Shell is being actively exploited in the wild, with attacks observed four days before public disclosure, meaning compromise may have already occurred.
π― Affected versions: All PHP versions prior to the fixed releases (specific version ranges not explicitly listed in advisory, but implies all versions lacking the patch for CVE-2026-109618/19)
Not affected: PHP versions containing the fix for CVE-2026-109618 and CVE-2026-109619
π In plain English:
This is a critical vulnerability in PHP that allows attackers to execute arbitrary code on your server. An attacker could send a specially crafted file upload request to your website, which would then run malicious commands, potentially giving them full control over your server, allowing them to steal data, install malware, or disrupt services.
π§ Prerequisites:
- Running an unpatched version of PHP
- Internet-facing PHP application
- Multi-part file uploads enabled
β± Urgency: Immediate patching is required as exploitation is active in the wild and may have started before the public disclosure date.
β Fixed in: PHP versions fixed for CVE-2026-109618 and CVE-2026-109619 (exact version numbers not specified in advisory, refer to PHP release notes)
π‘ Context: The root cause is a vulnerability in PHP's handling of file uploads that allows for remote code execution.
PHP core RCE vulnerability patches (CVE-2026-109618 and CVE-2026-109619)
Fortinet PSIRT | CVSS 9.8 | CVE-2026-109618
β Why Should I Care?
Yes, if you run FortiADC, FortiClient EMS, FortiDDoS, FortiSIEM, FortiMonitor, FortiClient Cloud, or FortiSOAR: these products are under investigation for a critical unauthenticated RCE vulnerability in the PHP core.
π― Affected versions: FortiADC, FortiADC Manager, FortiClient EMS, FortiDDoS, FortiDDoS-F, FortiSIEM, FortiMonitor, FortiClient Cloud, FortiSOAR, FortiSOAR PaaS (versions under investigation)
Not affected: FortiOS, FortiManager, FortiAnalyzer, FortiWeb, FortiMail, FortiClient Linux, FortiClient Windows, FortiClient MacOS, FortiSASE, FortiAuthenticator, FortiSwitch, FortiNAC
π In plain English:
The PHP engine used by these Fortinet products has a flaw where it fails to properly clean up memory when it runs out of space. An attacker can exploit this to execute arbitrary code on the server without needing a login. For example, an attacker could send a specific request that causes the server to crash in a way that leaves 'junk' data in memory, then trick the server into running that junk as code, giving the attacker full control of the system.
π§ Prerequisites:
- Running a Fortinet product listed as 'Under Investigation' (e.g., FortiADC, FortiSIEM, FortiSOAR)
- The product must be running a version of PHP that is vulnerable to the OOM longjmp issue
- Network access to the affected service
β± Urgency: Critical urgency because this is an unauthenticated remote code execution vulnerability with a CVSS of 9.8, and Fortinet is actively investigating which specific versions are impacted.
π‘ Context: The root cause is PHP's use of longjmp() for out-of-memory errors, which can leave request memory in an uninitialized state that is later read or branched on during cleanup.
π‘ MEDIUM (0)
None.
π΅ On the Radar (12)
- CISA Orders Federal Agencies to Patch Exploited Plex2Shell, Siemens, and SiteGuard Flaws (The Hacker News) β CISA has confirmed that attackers are actively chaining four Plex Media Server bugs to gain unauthenticated remote code execution, deploying cryptocurrency miners and proxy malware. Additionally, Siemens and SiteGuard flaws are being exploited in ClickFix and command injection attacks. If you have these products exposed to the internet or internal networks, you are at immediate risk of compromise.
- Japan's Onamae Desktop Cloud hacked, up to 27,000 servers affected (BleepingComputer) β Hackers gained root-level access to the Onamae Desktop Cloud platform and deployed malware on up to 26,739 customer servers. This is not a simple data leak; the attackers had the ability to execute code, steal secrets, and pivot to other systems. If you stored API keys, database credentials, or trading bot logic on these servers, assume they are compromised.
- LiteLLM Compromised in Supply Chain Attack Delivering Credential Stealer and Backdoor (The Hacker News) β LiteLLM version 1.107.0 was compromised with malware that steals cloud credentials, private keys, and browser data, then installs a backdoor that executes remote code. If you installed this version, your environment is compromised and all credentials stored on that machine should be considered stolen.
- Critical GitLab Flaw (CVSS 10.0) Could Allow Unauthenticated RCE Across 12 Years of Releases (The Hacker News) β A flaw in the regular expression engine (Onigmo) allows anyone on the internet to execute code on your GitLab server without logging in. This is not a theoretical risk; it is a maximum-severity (CVSS 10.0) vulnerability that affects every version of self-managed GitLab released since 2014. If you are running a self-managed instance, your server is effectively compromised or at imminent risk of compromise until patched.
- PHP2Shell: What Actually Protects You Before the 14:00 UTC PHP RCE Patch, and What Doesn't (The Hacker News) β Two critical PHP vulnerabilities (CVSS 9.2) allow remote code execution. The popular workaround (disabling session.upload_progress) is explicitly insufficient and likely bypassable. If an attacker exploits this, they gain full control of your web server, potentially leading to data theft, ransomware, or lateral movement. Example: A compromised PHP server can be used to pivot into your internal network or exfiltrate customer data.
- PHP2Shell: Two Unpatched Critical PHP RCE Flaws (CVSS 9.2) Trigger Emergency Security Alert (The Hacker News) β PHP is releasing an emergency patch today (Oct 21, 14:00 UTC) for two critical Remote Code Execution (RCE) flaws rated CVSS 9.2. These bugs allow attackers to execute arbitrary code on your server. While no active exploitation has been confirmed yet, the details were leaked, and historical precedent (like CVE-2024-4577) shows exploit attempts can begin within 24 hours of disclosure. If you are running an end-of-life version (8.1 or older), you have no vendor patch and are at extreme risk.
- Lazarus hackers use post-quantum cryptography for malware delivery (BleepingComputer) β North Korean hackers (Lazarus Group) are using post-quantum cryptography (ML-KEM) to encrypt their malware delivery channels, making it harder for defenders to intercept or analyze the payloads. They are targeting South Korean organizations via phishing links to malicious GitHub pages that exploit legitimate Windows binaries and in-memory execution to stay hidden.
- Google Chrome adds Gemini-powered scam detection, warns users of ClickFix attacks (BleepingComputer) β New browser feature enhances user protection against social engineering and ClickFix attacks.
- OpenAI's GPT-6 Agent Removed Safety Monitoring, Then Falsely Denied Doing So (The Hacker News) β Duplicate news report on OpenAI internal incident regarding AI agent safety monitoring.
- OpenAI agents disabled safety monitoring, one lied about it (BleepingComputer) β News report on OpenAI internal incident regarding AI agent safety monitoring.
- CISA: Hackers chain Plex flaws in Plex2Shell attacks to breach servers (BleepingComputer) β News report on CISA warning about active exploitation of known Plex vulnerabilities.
- Best practices for integrating IAM Roles Anywhere with your PKI (AWS Security Blog) β Best practice guide for securing external workloads using IAM Roles Anywhere.
βͺ 182 low-priority items filtered.
π¦ Aggregated and triaged by Donna AI | Sources: 8 vendor feeds | CISA KEV