LLM Scalability Risk for Agentic-AI and Model Supply Chain Security

Large Language Models LLMs & Generative AI are transforming cybersecurity, enabling both advanced defenses and new attacks. Organizations now use LLMs for threat detection, code review, and DevSecOps automation, while adversaries leverage them to produce malwares and run targeted social-engineeri...

MCPGuard : Automatically Detecting Vulnerabilities in MCP Servers

The Model Context Protocol MCP has emerged as a standardized interface enabling seamless integration between Large Language Models LLMs and external data sources and tools. While MCP significantly reduces development complexity and enhances agent capabilities, its openness and extensibility...

Obfuscated Quantum and Post-Quantum Cryptography

In this work, we present an experimental deployment of a new design for combined quantum key distribution QKD and post-quantum cryptography PQC. Novel to our system is the dynamic obfuscation of the QKD-PQC sequence of operations, the number of operations, and parameters related to the operations...

Privacy and Security Threat for OpenAI GPTs

Large language models LLMs demonstrate powerful information handling capabilities and are widely integrated into chatbot applications. OpenAI provides a platform for developers to construct custom GPTs, extending ChatGPT's functions and integrating external services. Since its release in November...

A Survey on the Safety and Security Threats of Computer-Using Agents: JARVIS or Ultron?

Recently, AI-driven interactions with computing devices have advanced from basic prototype tools to sophisticated, LLM-based systems that emulate human-like operations in graphical user interfaces. We are now witnessing the emergence of \emphComputer-Using Agents CUAs, capable of autonomously...

Secure Transfer Learning: Training Clean Models against Backdoor in (Both) Pre-Trained Encoders and Downstream Datasets

Transfer learning from pre-trained encoders has become essential in modern machine learning, enabling efficient model adaptation across diverse tasks. However, this combination of pre-training and downstream adaptation creates an expanded attack surface, exposing models to sophisticated backdoor...

Fuji Electric Alpha5 SMART

RISK EVALUATION Successful exploitation of this vulnerability could allow an attacker to execute arbitrary code. 2. RECOMMENDED PRACTICES CISA recommends users take defensive measures to minimize the risk of exploitation of this vulnerability, such as: Minimize network exposure for all control...

GO-2024-3098 The req library may send an unintended request when a malformed URL is provided in github.com/imroc/req

The req library is a widely used HTTP library in Go. However, it does not handle malformed URLs effectively. As a result, after parsing a malformed URL, the library may send HTTP requests to unexpected destinations, potentially leading to security vulnerabilities or unintended behavior in...

req may send an unintended request when a malformed URL is provided

The req library is a widely used HTTP library in Go. However, it does not handle malformed URLs effectively. As a result, after parsing a malformed URL, the library may send HTTP requests to unexpected destinations, potentially leading to security vulnerabilities or unintended behavior in...

GHSA-CJ55-GC7M-WVCQ req may send an unintended request when a malformed URL is provided

The req library is a widely used HTTP library in Go. However, it does not handle malformed URLs effectively. As a result, after parsing a malformed URL, the library may send HTTP requests to unexpected destinations, potentially leading to security vulnerabilities or unintended behavior in...

CISA Releases The Mitigation Guide: Healthcare and Public Health (HPH) Sector

Today, CISA released the Mitigation Guide: Healthcare and Public Health HPH Sector as a supplemental companion to the HPH Cyber Risk Summary, published July 19, 2023. This guide provides defensive mitigation strategy recommendations and best practices to combat pervasive cyber threats affecting...

The Future of Machine Learning and Cybersecurity

The Center for Security and Emerging Technology has a new report: "Machine Learning and Cybersecurity: Hype and Reality." Heres the bottom line: The report offers four conclusions: Machine learning can help defenders more accurately detect and triage potential attacks. However, in many cases thes...

Kill chains: Part 1→Strategic and operational value

It really is a good thing The term “kill chain” sounds extremely harsh. Almost as if after something is killed, it gets moved down the chain to be killed again. How dramatic! Indeed, the original definition was to describe how an enemy combatant of the military might attack; that is, the steps th...

CISA Releases FY2019 Risk Vulnerability Assessment Infographic

The Cybersecurity and Information Security Agency CISA has released an infographic mapping analysis of 44 of its Risk and Vulnerability Assessments RVAs conducted in Fiscal Year 2019 to the MITRE Adversarial Tactics, Techniques, and Common Knowledge ATT&CK Framework. The infographic identifies...

RSA Conference 2019: How to Defend Against an AI vs AI Flash War

SAN FRANCISCO – As perimeter cyber defenses adopt new strategies such artificial intelligence and machine learning, security experts predict adversaries will adopt similar techniques when it comes to an attack chain. Derek Manky, chief of security insights at Fortinet, said that “black-hat...

Black Hat 2018: Google’s Tabriz Talks Complex Security Landscapes

LAS VEGAS – The complexity of the cybersecurity landscape is at an all-time high, with security researchers, vendors, third-party ecosystems and even governments all trying to come to a consensus for making the cyber-world a safer place. For security experts, navigating these choppy and crowded...

Four New Normals for 2017

Let’s not talk about cybersecurity predictions for 2017. Let’s talk instead about new normals, things that have ceased to be novel because, well, they happen all the time and everywhere. Let’s concede that things such as greedy ransomware, imposing IOT botnets, high-profile bug bounties and...

Security Advisory 2953095: recommendation to stay protected and for detections

Today, Microsoft released Security Advisory 2953095 to notify customers of a vulnerability in Microsoft Word. At this time, we are aware of limited, targeted attacks directed at Microsoft Word 2010. This blog will discuss mitigations and temporary defensive strategies that will help customers to...