Tag: nvidia
32 articles
Nvidia Bug Exposes Thousands of GPU Servers to Potential Disruption
A newly discovered vulnerability in Nvidia's DCGM Exporter, tracked as CVE-2026-47483, has left around 2,100 GPU servers exposed to potential disruption, with a staggering 12,000 GPU UUIDs and $100M in hardware at risk. Researchers warn that unauthenticated actors could exploit this flaw to crash the GPU monitoring service, making a swift patch to version 4.8.2 or later a top priority.
US Charges Californian with Smuggling $300M in Nvidia Chips to China
Meet Greg Lui, a Californian who's in hot water for allegedly smuggling $300 million worth of high-end Nvidia chips to China, dodging US export licenses in the process. He's facing up to 50 years in prison for his role in the scheme, which allegedly ran from 2023 to 2026.
AI Executives Sign Accord Amid Rising Cybersecurity Concerns
In a landmark move, President Donald Trump and top AI executives from OpenAI, Anthropic, Google, Meta, Nvidia, and xAI have signed a voluntary safety accord, pledging to enact multiple layers of controls to tackle rising cybersecurity concerns. This non-binding agreement, described as "morally binding" and "almost like a constitution," aims to prioritize safety and security in AI development.
NVIDIA Bolsters AI Security with Open-Source Agent Safety Platform
NVIDIA just unveiled an open-source platform to ensure AI agents operate safely and transparently, earning trust through robust control and governance. The Open Agent Safety Platform has already garnered support from over 100 AI organizations.
NVIDIA Unveils Platform to Secure Autonomous AI Agents
NVIDIA's new Open Agent Safety Platform is a game-changer, bringing enforceable controls to autonomous AI systems to prevent them from going off the rails - and over 100 top organizations, including Microsoft and Salesforce, are already on board. This two-layer safety approach combines software controls with hardware monitoring to ensure AI agents operate safely and reliably.
NVIDIA GPUs Vulnerable to GPUThor Rowhammer Attack Bypassing ECC
NVIDIA GPUs are vulnerable to a new type of attack, dubbed GPUThor, which can cause massive corruption - up to 377,552 bit flips per gigabyte - on certain models, including the RTX A5000, when their GDDR6 memory is exploited. This Rowhammer-style attack can bypass ECC protection, putting powerful NVIDIA workstation cards at risk.
GPUThor Attack Bypasses NVIDIA ECC Protection
Meet GPUThor, a game-changing attack that shatters NVIDIA's ECC protection, making it alarmingly easy to execute practical root-level attacks. This sinister technique can trigger a staggering 72,000 to 377,000 bit flips per gigabyte, putting even the toughest defenses to shame.
NVIDIA NemoClaw Exposes AI Models to Poisoning via Webpage
NVIDIA's NemoClaw exposes AI models to poisoning via webpage, allowing attackers to take control of the model server by binding it to every network interface. This configuration vulnerability makes it easy for hackers to access and manipulate the Ollama API from outside the loopback address.
Tech Giants Propose SAFE Framework for AI Threat Intel Sharing
The tech community is stepping up to tackle AI threats with a new framework called SAFE, designed to facilitate confidential sharing of AI operational failures and translate those lessons into actionable defensive guidance. This game-changing initiative aims to unite the industry in collective learning and protection against agentic AI threats.
Tech Giants Form Alliance to Bolster Open-Source AI Models
In a bid to revolutionize AI security, 37 tech giants, led by NVIDIA, have joined forces to form the Open Secure AI Alliance, championing the development of open-source AI models that boost cybersecurity and foster customization. By uniting behind open models, harness, and tools, the alliance aims to safeguard infrastructure and empower defenders to adapt and deploy robust protections.
NVIDIA Launches Open Secure AI Alliance Without Key Players
NVIDIA is shaking up the AI security landscape with the launch of the Open Secure AI Alliance, a groundbreaking coalition of nearly 40 tech giants, including Adobe, Cisco, and Microsoft, dedicated to developing open-source security tools to safeguard artificial intelligence. By joining forces, they're building a robust defense stack to protect AI agents from identity threats to secure coding workflows.
NVIDIA Launches Open Secure AI Alliance to Share Threat-Detecting Tech
Join the Open Secure AI Alliance, a groundbreaking coalition of 37 industry leaders, as they revolutionize AI security by sharing cutting-edge threat-detecting technologies and collaborative defense strategies. Together, they're breaking down silos to safeguard the future of AI and software development.
Nvidia Launches Open Secure AI Alliance to Promote Open-Source Models
Nvidia has launched the Open Secure AI Alliance, a groundbreaking coalition with industry giants like Microsoft, IBM, and Adobe, to revolutionize national cyber defenses with open-source AI models that are trustworthy, transparent, and controllable. By joining forces, these leaders aim to empower defenders worldwide with cutting-edge, open tools to stay ahead of emerging threats.
LabubaRAT Exploits NVIDIA Disguise to Control Windows Hosts
Meet LabubaRAT, a sneaky threat that masquerades as NVIDIA software to take control of Windows hosts, allowing hackers to profile, capture, and manipulate sensitive data. Once deployed, it creates a hidden backdoor for further malicious activity.
Linux Rootkits Persist in Updated Forms
A single misstep with an over-privileged or poorly designed agent can quickly spiral into a serious incident, making the UK National Cyber Security Centre's warning feel alarmingly relevant. This urgency was underscored at Pwn2Own Berlin, where researchers exploited 47 zero-day flaws, raking in over $1.2 million in rewards.
US Approves AI Chip Sales to China, Sparking Security Fears
The US has greenlit the sale of AI chips to China, with around 10 Chinese firms cleared to buy Nvidia's H200 chips, sparking concerns about national security and the rapidly evolving AI landscape. Treasury Secretary Scott Bessent hinted at upcoming discussions with China on implementing AI safeguards.
Foxconn Cyberattack Exposes Sensitive Data from Apple, Nvidia Projects
A massive cyberattack on Foxconn's North American factories has compromised sensitive data from major tech giants, including Apple and Nvidia, with hackers allegedly making off with a staggering 8 TB of data and over 11 million files. Foxconn has confirmed the breach, assuring that production is resuming after swiftly activating its cybersecurity response mechanism.
NVIDIA Discloses GeForce NOW Breach Affecting Armenian Users
NVIDIA recently discovered a security breach affecting users of GeForce NOW in Armenia, which was caused by a compromised system operated by a third-party partner, not by NVIDIA's own network. The company is working closely with the partner to resolve the issue and notify affected users.
MD5 Password Hashes Cracked in Under an Hour
In a shocking test, Kaspersky researchers cracked 60% of 231 million MD5 password hashes in under an hour using just one high-powered graphics card, revealing the alarming vulnerability of even the most seemingly secure passwords. This unsettling experiment highlights the urgent need for stronger password protection.
NVIDIA Chips Vulnerable to Rowhammer Attacks
Researchers have discovered that NVIDIA chips are vulnerable to Rowhammer attacks, which can be exploited to gain unauthorized access to computer systems. This security threat can lead to a complete compromise of the machine, allowing attackers to read and write data freely.
Pentagon Clears 7 AI Firms for Classified Network Access
The Pentagon has cleared seven top AI firms, including Amazon, Google, and Microsoft, to access classified Defense Department networks, marking a major step forward in harnessing AI for national security. This strategic move aims to prevent reliance on a single vendor and drive American leadership in AI.
Nvidia Launches AI Models to Tame Quantum Computing's Error Challenge
Nvidia is harnessing the power of artificial intelligence to tackle the error challenge in quantum computing, introducing the Ising Model Family, a set of open AI models designed to reduce errors and make noisy quantum hardware more useful. By pairing smarter software with existing quantum machines, Nvidia aims to bring us closer to unlocking the full potential of quantum computing.
Popular Python libraries: Stunning Hugging Face danger
Think twice before blindly loading Hugging Face models: researchers found attackers can hide executable Python code in file metadata and malformed pickles so a downloaded model can automatically run malicious payloads. With major libraries and millions of downloads affected, this stealthy supply‑chain trick puts countless projects and machines at risk.
AI Bugs: Stunning Critical Flaws Expose Meta, Nvidia, MS
Turns out the plumbing behind the AI boom—ZMQ messaging and unsafe pickle deserialization—can let attackers slip in and execute arbitrary code. Recent disclosures expose critical remote‑code vulnerabilities across Meta, Nvidia, Microsoft and popular PyTorch inference stacks.