📄 الوصف (الإنجليزية)
A vulnerability has been found in ChatGPTNextWeb NextChat up to 2.16.1. Affected is the function addMcpServer of the file app/mcp/actions.ts. The manipulation leads to improper authorization. Remote exploitation of the attack is possible. The exploit has been disclosed to the public and may be used. The project was informed of the problem early through an issue report but has not responded yet.
🤖 ملخص AI
CVE-2026-7644 is a high-severity improper authorization vulnerability in ChatGPTNextWeb NextChat versions up to 2.16.1 affecting the MCP server addition function. The vulnerability allows remote exploitation without authentication bypass, potentially enabling unauthorized users to add malicious MCP servers to the application. With no patch currently available and public exploit disclosure, this poses an immediate risk to organizations using affected versions.
📄 الوصف (العربية)
🤖 التحليل الذكي آخر تحليل: May 6, 2026 20:01
🇸🇦 التأثير على المملكة العربية السعودية
Saudi organizations using ChatGPTNextWeb NextChat for AI-powered customer service, internal knowledge management, or research purposes face significant risk. Most vulnerable sectors include: Banking and Financial Services (SAMA-regulated entities using AI chatbots for customer interaction), Government agencies (NCA oversight, digital transformation initiatives), Healthcare providers (patient information systems), and Telecommunications (STC, Mobily customer support systems). The improper authorization flaw could allow threat actors to inject malicious MCP servers, leading to data exfiltration, prompt injection attacks, or lateral movement within enterprise networks. Organizations in critical infrastructure sectors are particularly at risk due to potential supply chain compromise.
🏢 القطاعات السعودية المتأثرة
Banking and Financial Services
Government and Public Administration
Healthcare and Medical Services
Energy and Utilities
Telecommunications
Education and Research
E-commerce and Retail
🎯 تقنيات MITRE ATT&CK
T1078 - Valid Accounts (improper authorization allows unauthorized access)
T1190 - Exploit Public-Facing Application (remote exploitation possible)
T1199 - Trusted Relationship (MCP server injection for supply chain compromise)
T1021 - Remote Services (lateral movement via compromised MCP servers)
T1566 - Phishing (malicious MCP server injection for credential harvesting)
T1557 - Man-in-the-Middle (MCP server interception)
T1040 - Traffic Capture (data exfiltration via MCP channels)
⚖️ درجة المخاطر السعودية (AI)
7.8
/ 10.0
🔧 Remediation Steps (English)
Immediate Actions:
1. Inventory all instances of ChatGPTNextWeb NextChat running versions ≤2.16.1 across your organization
2. Restrict network access to affected instances using firewall rules and WAF policies
3. Implement strict input validation on MCP server addition endpoints
4. Enable comprehensive logging and monitoring of MCP server configuration changes
5. Review audit logs for unauthorized MCP server additions or suspicious configuration modifications
Compensating Controls (until patch available):
6. Implement network segmentation to isolate ChatGPT instances from sensitive systems
7. Deploy API gateway with authentication enforcement for MCP server endpoints
8. Use role-based access control (RBAC) to restrict MCP server management to authorized administrators only
9. Implement request signing and mutual TLS for MCP server communications
10. Monitor for suspicious MCP server connections and data flows
Detection Rules:
11. Alert on any POST/PUT requests to /app/mcp/actions.ts addMcpServer endpoint from non-whitelisted IPs
12. Monitor for MCP server configuration changes outside maintenance windows
13. Track failed authentication attempts to MCP management interfaces
14. Flag connections to unknown or suspicious MCP servers
Long-term:
15. Plan immediate upgrade to patched version once available (monitor GitHub releases)
16. Consider alternative AI solutions with stronger authorization controls if patch delays exceed 30 days
🔧 خطوات المعالجة (العربية)
الإجراءات الفورية:
1. قم بحصر جميع نسخ ChatGPTNextWeb NextChat التي تعمل بالإصدارات ≤2.16.1 عبر مؤسستك
2. قيّد الوصول إلى الشبكة للنسخ المتأثرة باستخدام قواعد جدار الحماية وسياسات WAF
3. طبّق التحقق الصارم من صحة المدخلات على نقاط نهاية إضافة خادم MCP
4. فعّل السجلات الشاملة ومراقبة تغييرات تكوين خادم MCP
5. راجع سجلات التدقيق للتحقق من إضافات خادم MCP غير المصرح بها أو التعديلات المريبة
الضوابط البديلة (حتى توفر التصحيح):
6. طبّق تقسيم الشبكة لعزل نسخ ChatGPT عن الأنظمة الحساسة
7. نشّر بوابة API مع فرض المصادقة لنقاط نهاية خادم MCP
8. استخدم التحكم في الوصول القائم على الأدوار (RBAC) لتقييد إدارة خادم MCP للمسؤولين المصرح لهم فقط
9. طبّق توقيع الطلب و TLS المتبادل لاتصالات خادم MCP
10. راقب اتصالات خادم MCP المريبة وتدفقات البيانات
قواعد الكشف:
11. أصدر تنبيهات لأي طلبات POST/PUT إلى نقطة نهاية addMcpServer من عناوين IP غير مدرجة في القائمة البيضاء
12. راقب تغييرات تكوين خادم MCP خارج نوافذ الصيانة
13. تتبع محاولات المصادقة الفاشلة لواجهات إدارة MCP
14. علّم الاتصالات بخوادم MCP غير معروفة أو مريبة
المدى الطويل:
15. خطّط للترقية الفورية إلى الإصدار المصحح بمجرد توفره (راقب إصدارات GitHub)
16. فكّر في حلول ذكاء اصطناعي بديلة بضوابط تفويض أقوى إذا تجاوزت تأخيرات التصحيح 30 يوماً
📋 خريطة الامتثال التنظيمي
🟢 NCA ECC 2024
ECC 2024 A.5.1.1 - Access Control Policy (improper authorization violates access control principles)
ECC 2024 A.5.2.1 - User Registration and Access Rights Management
ECC 2024 A.5.3.1 - Management of Privileged Access Rights
ECC 2024 A.8.2.1 - User Access Management
ECC 2024 A.9.2.1 - User Access Rights Review
🔵 SAMA CSF
SAMA CSF ID.AC-1 - Access Control Policy and Procedures
SAMA CSF PR.AC-1 - Identities and Credentials are Managed
SAMA CSF PR.AC-3 - Access Restrictions and Management of Physical and Logical Assets
SAMA CSF DE.CM-1 - The Network is Monitored to Detect Potential Cybersecurity Events
🟡 ISO 27001:2022
ISO 27001:2022 A.5.2 - Information Security Policies
ISO 27001:2022 A.8.2 - Privileged Access Rights
ISO 27001:2022 A.8.3 - Information Access Restriction
ISO 27001:2022 A.9.2 - User Access Management
ISO 27001:2022 A.9.4 - Access Rights Review
🟣 PCI DSS v4.0.1
PCI DSS 2.1 - Establish Configuration Standards
PCI DSS 7.1 - Limit Access to System Components
PCI DSS 8.1 - Assign Unique ID to Each User
PCI DSS 8.2 - Ensure User Identity is Verified Before Modifying Authentication Credentials