admin

admin

  • Joined: September 10, 2024
  • Articles: 5734

Security Flaws in NVIDIA Triton Allow Unauthenticated Attacks to Execute Code and Compromise AI Servers

Published: August 4, 2025
Category: AI Security / Vulnerability

A newly revealed set of vulnerabilities in NVIDIA’s Triton Inference Server—an open-source platform for deploying artificial intelligence (AI) models across Windows and Linux—puts susceptible servers at risk of takeover. Researchers Ronen Shustin and Nir Ohfeld from Wiz noted in a report released today that when these flaws are exploited together, they could enable remote, unauthenticated attackers to gain full control of the server, facilitating remote code execution (RCE). The identified vulnerabilities include:

  • CVE-2025-23319 (CVSS Score: 8.1): An issue in the Python backend that allows for an out-of-bounds write via specifically crafted requests.
  • CVE-2025-23320 (CVSS Score: 7.5): A flaw in the Python backend where an attacker can exceed the shared memory limit by sending an excessively large request.
  • CVE-2025-23334 (CVSS Score: 5.9): A vulnerability in the Python backend that could lead to an out-of-bounds read.

NVIDIA Triton Vulnerabilities Enable Unauthenticated Code Execution Risks in AI Servers August 4, 2025 A critical security issue has emerged concerning NVIDIA’s Triton Inference Server, a widely used open-source platform designed for deploying artificial intelligence models on Windows and Linux…

Fortinet Issues Critical Patch for SQL Injection Vulnerability in FortiWeb (CVE-2025-25257)

July 11, 2025, United States

Fortinet has unveiled a patch addressing a severe security vulnerability in FortiWeb, which could allow unauthenticated attackers to execute arbitrary database commands on affected systems. Designated as CVE-2025-25257, this flaw has a CVSS score of 9.6 out of 10. According to Fortinet’s advisory, the vulnerability stems from “improper neutralization of special elements used in an SQL command (SQL Injection) [CWE-89],” enabling unauthorized SQL code execution through specially crafted HTTP or HTTPS requests.

The vulnerability affects the following FortiWeb versions:

  • FortiWeb 7.6.0 to 7.6.3 (Upgrade to 7.6.4 or higher)
  • FortiWeb 7.4.0 to 7.4.7 (Upgrade to 7.4.8 or higher)
  • FortiWeb 7.2.0 to 7.2.10 (Upgrade to 7.2.11 or higher)
  • FortiWeb 7.0.0 to 7.0.10 (Upgrade to 7.0.11 or higher)

Kentaro Kawane from GMO Cybersecurity is credited with reporting this significant vulnerability, as well as several critical issues in Cisco systems.

Fortinet Issues Critical Patch for SQL Injection Vulnerability in FortiWeb On July 11, 2025, Fortinet announced the release of urgent patches for a significant security vulnerability in FortiWeb, a web application firewall. This flaw, designated CVE-2025-25257, poses a serious risk,…

GPUHammer: New RowHammer Attack Variant Compromises AI Model Integrity on NVIDIA GPUs

NVIDIA is advising customers to activate System-level Error Correction Codes (ECC) as a safeguard against a newly identified variant of the RowHammer attack targeting its graphics processing units (GPUs). “The likelihood of successful RowHammer exploitation varies depending on DRAM device, platform, design specifications, and system settings,” the company noted in a recent advisory. Named GPUHammer, this marks the first incident of a RowHammer exploit impacting NVIDIA GPUs, such as the A6000 with GDDR6 memory. This attack allows malicious users to manipulate other users’ data by inducing bit flips in GPU memory. Researchers from the University of Toronto highlighted a particularly alarming outcome: the accuracy of an AI model can plummet from 80% to below 1%. RowHammer poses a similar risk to modern DRAMs as Spectre and Meltdown do for contemporary CPUs, representing critical hardware-level security vulnerabilities.

GPUHammer: New RowHammer Attack Variant Threatens AI Performance on NVIDIA GPUs On July 12, 2025, NVIDIA issued a critical advisory urging its customers to activate System-level Error Correction Codes (ECC) to combat a newly revealed variant of RowHammer attacks targeting…

New GootLoader Campaign Targets Those Searching for Bengal Cat Regulations in Australia

Date: Nov 11, 2024
Category: Malware / SEO Poisoning

In a uniquely targeted effort, individuals looking for information on the legality of Bengal Cats in Australia are falling victim to the GootLoader malware. “We discovered GootLoader operators utilizing search inquiries regarding a specific cat breed and region to deliver malware: ‘Are Bengal Cats legal in Australia?'” noted Sophos researchers Trang Tang, Hikaru Koike, Asha Castle, and Sean Gallagher in a report released last week. GootLoader, as its name suggests, is a malware loader typically spread through search engine optimization (SEO) poisoning techniques for initial entry. The malware is triggered when users search for terms related to legal documents and agreements; this leads to compromised links that direct them to infected websites hosting a ZIP file containing a JavaScript payload. Once executed, it paves the way for further malicious software installation.

New GootLoader Campaign Targets Searches for Bengal Cat Laws in Australia In a targeted cybersecurity threat, attackers are leveraging interest in the legality of Bengal cats in Australia to distribute GootLoader malware. This specific campaign highlights the methodical approach employed…