Researchers released a detailed report on a significant security vulnerability named CVE-2023-49785, affecting the ChatGPT Next Web, popularly known as NextChat.
This vulnerability has raised concerns within the cybersecurity community due to its potential for exploitation through Server-Side Request Forgery (SSRF).
NextChat is a web interface designed for large language model (LLM) services. It provides a platform for users to interact with advanced AI models like ChatGPT.
It facilitates seamless integration and functionality across various platforms, making it a widely used tool for businesses and individuals alike.
Free Webinar on Best Practices for API vulnerability & Penetration Testing: Free Registration
The vulnerability, identified as CVE-2023-49785, was publicly disclosed on March 11, 2024. It affects versions of NextChat 2.11.2 and earlier, as per a report by the S2W Threat Intelligence Center.
The issue was traced to the API responsible for client settings synchronization, which lacked appropriate security measures, allowing unauthorized users to exploit the system.
The SSRF vulnerability stemmed from the improperly secured/API/cors endpoint. This endpoint circumvents CORS policies by routing requests through the server.
Unfortunately, it was accessible to unauthenticated users, who could manipulate it to send malicious requests to arbitrary internal services.
Exploiting this vulnerability, attackers could perform SSRF actions to gain unauthorized access to sensitive server information.
This breach could lead to the theft of authentication data or the misuse of server privileges, potentially compromising entire cloud environments.
To address this critical flaw, users of NextChat are advised to take immediate action:
The full report, which offers an in-depth analysis and further recommendations, is available for security teams and developers who wish to secure their NextChat deployments fully.
This incident underscores the importance of regular security assessments and timely updates to safeguard against evolving threats in the digital landscape.
Analyse Real-World Malware & Phishing Attacks With ANY.RUN - Get up to 3 Free Licenses
The eSentire’s Threat Response Unit (TRU) in early March 2025, a sophisticated cyberattack leveraging SocGholish…
The NSFOCUS Fuying Laboratory’s global threat hunting system identified 19 sophisticated Advanced Persistent Threat (APT)…
The FBI’s Internet Crime Complaint Center (IC3) has reported a record-breaking loss of $16.6 billion…
Cybersecurity researchers from The DFIR Report’s Threat Intel Group uncovered an open directory hosted at…
A newly analyzed Python-based Remote Access Trojan (RAT) has emerged as a significant cybersecurity threat,…
A multi-stage carding attack has been uncovered targeting a Magento eCommerce website running an outdated…