DARPA Picks Seven Small Businesses for AI Cyber Challenge

The Defense Advanced Research Projects Agency (DARPA) has awarded seven small businesses contracts worth $1 million each to develop a cyber reasoning system using artificial intelligence (AI).

Researchers Warn of Stuxnet-Style Web-Based PLC Malware

Researchers from the Georgia Institute of Technology have developed web-based malware called IronSpider, targeting modern programmable logic controllers (PLCs) used in industrial control systems.

Self-Propagating Worm Created to Target Generative AI Systems

Researchers from Israel Institute of Technology, Intuit and Cornell Tech have developed a computer worm called "Morris II" that targets generative AI (GenAI) applications to spread malware and steal personal data.

BEAST Attack on AI Models can Break LLM Guardrails in a Minute

Computer scientists have developed a fast and efficient method, called BEAST, to generate harmful prompts that elicit undesirable responses from large language models using an Nvidia RTX A6000 GPU with 48GB of memory.

PayPal Files Patent for New Method to Detect Stolen Cookies

PayPal has filed a patent application for a method to detect when "super-cookies" are stolen, aiming to improve cookie-based authentication and prevent account takeover attacks.

Hackers can Use Generative AI to Manipulate Live Conversations

IBM researchers demonstrated a technique to intercept live conversations and replace keywords based on the context, allowing for the manipulation of information, financial fraud, and even real-time changes to news broadcasts and political speeches.

New Method To Safeguard Against Mobile Account Takeovers

The method involves modeling how account access changes as devices, SIM cards, or apps are disconnected from the account ecosystem, providing insights into complex hacking attacks.

Researchers Develop Technique to Prevent Software Bugs

A team of computer scientists has developed a method called Baldur, which uses artificial intelligence to automatically generate proofs and verify the correctness of software, aiming to reduce software bugs and vulnerabilities.

Researchers Automated Jailbreaking of LLMs With Other LLMs

Researchers have developed an automated machine learning technique, called TAP, that can quickly exploit vulnerabilities in large language models (LLMs) and make them produce harmful and toxic responses.

New BLUFFS Attack Lets Attackers Hijack Bluetooth Connections

Considering the widespread use of the well-established wireless communication standard and the versions impacted by the exploits, BLUFFS could work against billions of devices, including laptops, smartphones, and other mobile devices.

Defend Against Threats with Cyber Fusion

Cyware is the leading provider of cyber fusion solutions that power threat intelligence sharing , end-to-end automation and 360-degree threat response.

Trending Tags