AI & ML Security

An unprecedented surge in agent-led activity has forced the RubyGems registry to suspend new user registrations to mitigate systemic supply chain risks. The event, which unfolded over a 48-hour window in May, saw the injection of more than 2,000 malicious packages into the public repository,

The phenomenon of model theft allows competitors to extract the reasoning capabilities and features of frontier models through persistent inference attacks. This shift marks a departure from the days when cyber defense relied almost exclusively on firewalls and file-level encryption. In the current

Cybersecurity researchers have identified a new illicit artificial intelligence service being marketed to malicious actors as an uncensored alternative to mainstream assistants. This tool, known as Luciferus, represents a significant escalation in the ongoing arms race between security

Project-level isolation has become a critical requirement for firms attempting to scale AI agent fleets without compromising organizational integrity. The rapid integration of autonomous systems into the corporate landscape has transitioned from a speculative technological trend to a strategic

The convergence of IT and OT networks has created a scenario where network congestion from a cyberattack can trigger physical malfunctions in heavy machinery. This reality marks a departure from the days when industrial plants were safely tucked away behind physical air gaps. In 2026, the

The recent discovery of the Luciferus platform on the Exploit forum reveals a growing trend where cybercriminals utilize sophisticated machine learning models to automate the generation of malicious code. This illicit service, identified by researchers from the Sophos Counter Threat Unit (CTU), is