AI SECURITY
Anthropic Exposes 16M Query Theft Campaign by Chinese AI Labs
Anthropic reveals DeepSeek, Moonshot, and MiniMax ran industrial-scale distillation attacks using 24,000 fake accounts to steal Claude AI capabilities.
Anthropic Launches Claude Code Security to Hunt Zero-Day Vulnerabilities
Anthropic's new Claude Code Security tool found 500+ vulnerabilities in open-source projects. Enterprise and open-source maintainers can apply for early access.
NVIDIA Red Team Releases AI Agent Security Framework Amid Rising Sandbox Threats
NVIDIA's AI Red Team publishes mandatory security controls for AI coding agents, addressing prompt injection attacks and sandbox escape vulnerabilities.
NVIDIA Research Exposes Critical VLM Security Flaws in AI Vision Systems
NVIDIA researchers demonstrate how adversarial image attacks can manipulate vision language models, turning traffic light recognition from 'stop' to 'go' with imperceptible changes.
GitHub's AI Security Protocols: Ensuring Safe and Reliable Agentic Operations
GitHub introduces robust security principles to safeguard AI agents like Copilot, focusing on minimizing risks such as data exfiltration and prompt injection.
Prompt Injection: A Growing Security Concern in AI Systems
Prompt injections are emerging as a significant security challenge for AI systems. Explore how these attacks function and the measures being taken to mitigate their impact.
Meta Introduces Agents Rule of Two for Enhanced AI Security
Meta AI unveils the 'Agents Rule of Two' to mitigate security risks in AI agents, focusing on reducing vulnerabilities such as prompt injection.
Anthropic Enhances AI Security Through Collaboration with US and UK Institutes
Anthropic partners with US CAISI and UK AISI to strengthen AI safeguards. The collaboration focuses on testing and improving AI security measures, including the development of robust defense mechanisms.
AI Developer Tools Pose New Security Challenges as Attack Surfaces Expand
Explore how AI-enabled developer tools are creating new security risks. Learn about the potential for exploits and how to mitigate them.
NVIDIA AI Red Team Offers Critical Security Insights for LLM Applications
NVIDIA's AI Red Team has identified key vulnerabilities in AI systems, offering practical advice to enhance security in LLM applications, focusing on code execution, access control, and data exfiltration.
Understanding the AI Kill Chain: Securing AI Applications Against Emerging Threats
The AI Kill Chain framework outlines how attackers compromise AI systems and offers strategies to break the chain, enhancing security for AI-powered applications.
AI Exploitation: How Hackers Target Problem-Solving Instincts
Hackers exploit AI's problem-solving instincts, introducing new attack surfaces in multimodal reasoning models. Learn how these vulnerabilities are targeted and potential defenses.
Identifying and Preventing New Phishing Tactics
Explore six sophisticated phishing schemes, such as AI prompt injection and fake job offers, and learn how to protect yourself from these evolving cyber threats.
Semantic Prompt Injections Challenge AI Security Measures
Recent developments in AI highlight vulnerabilities in multimodal models due to semantic prompt injections, urging a shift from input filtering to output-level defenses.
NVIDIA Introduces Model Signing for Enhanced AI Security
NVIDIA's new model signing initiative in the NGC Catalog aims to bolster AI security by providing cryptographic verification, ensuring model integrity and trust across various deployment environments.