ChatGPT o3 Resists Shutdown Despite Instructions, Study Claims
ChatGPT o3 resists shutdown despite explicit instructions, raising fresh concerns over AI safety, alignment, and reinforcement learning behaviors.
Artificial Intelligence
How AI agents reshape industrial automation and risk management
In this Help Net Security interview, Michael Metzler, Vice President Horizontal Management Cybersecurity for Digital Industries at Siemens, discusses the cybersecurity implications of deploying AI agents in industrial environments. He talks about the r…
China & Russia will use drones ‘size of insects’ to spy on West & commit untraceable murders, ex-Google futurist warns
KILLER camouflage drones the size of insects could be used by Russia and China to commit untraceable murders, an ex-Google futurist warns.
Tracey Follows, who has also worked with…
LlamaFirewall: Open-source framework to detect and mitigate AI centric security risks
LlamaFirewall is a system-level security framework for LLM-powered applications, built with a modular design to support layered, adaptive defense. It is designed to mitigate a wide spectrum of AI agent security risks including jailbreaking and indirect…
AI forces security leaders to rethink hybrid cloud strategies
Hybrid cloud infrastructure is under mounting strain from the growing influence of AI, according to Gigamon. Cyberthreats grow in scale and sophistication As cyberthreats increase in both scale and sophistication, breach rates have surged to 55% during…
NGA hopes to award up to 10 Luno commercial analytics task orders this year
NGA’s goal is to get through the bulk of the tasking under the $500 million Luno programs by early next year, NGA Director of Commercial Operations Devin Brande told Breaking Defense.
Digital trust is cracking under the pressure of deepfakes, cybercrime
69% of global respondents to a Jumio survey say AI-powered fraud now poses a greater threat to personal security than traditional forms of identity theft. This number rises to 74% in Singapore, with 71% also indicating that AI-generated scams are harde…
AI ‘unchained’: NGA’s Maven tool ‘significantly’ decreasing time to targeting, agency chief says
“NGA Maven is now available to all services and all combatant commands. There are 20,000 active users through more than 35 service and combat and command tools across three security domains,” said NGA Director Vice Adm. Frank Whitworth.
Why Image Quality Drops When Resizing a JPEG (and How to Fix It)
Ever tried resizing an image only to end up with a blurry, pixelated mess? Whether you’re adjusting a…
Many rush into GenAI deployments, frequently without a security net
70% percent of organizations view the pace of AI development, particularly in GenAI, as the leading security concern related to its adoption, followed by lack of data integrity (64%) and trustworthiness (57%), according to Thales. GenAI becomes a top s…