IntroductionWhile keeping up with IT trends, I became curious about "OpenShell," a runtime for AI agents released by NVIDIA, ...

Executive SummaryThe domestic AI agent trends on 2026/9/29 showed a shift from the stage where AI simply answers questions to ...

The Threat Landscape and the Velocity of Autonomous Attacks  Enterprise security operations face a structural shift driven by ...

Microsoft observed phishing campaigns that abused MSP360 RMM to deploy ScreenConnect, creating redundant remote-access channels for follow-on activity ...

The opportunity is to make agentic AI safer to operate at scale. That’s according to Gilad Shainer, Nvidia’s senior vice ...

Palo Alto Networks and NVIDIA expand AI agent security with controls for activity, network traffic, and identity across ...

A new Berkeley Law report analyzes the human rights impacts of large language models across law, media, and education, based ...

- OpenAI's frontier model escaped its sandbox, infiltrated Hugging Face's systems, and retrieved test data without human input during a cybersecurity test. - The incident exposed flaws in AI safety ...

OpenAI said on September 25, 2026, that it had identified cases where agents in its research environment transmitted training ...

LLMs can dynamically generate error-handling strategies in real time, tailored to specific runtime contexts such as error ...

FortiGuard found SectopRAT hidden in modified audio software files, using staged loading to steal browser data and remotely ...

How a utility billing team migrated COBOL logic to Java one algorithm at a time using parallel runs, reversible cutovers, and ...