Enterprise Security Tech

NVIDIA has introduced the Open Agent Safety Platform, an open software platform and reference architecture intended to secure autonomous AI systems from testing through deployment. The initiative is designed to connect controls across the agent runtime, underlying compute infrastructure and physical systems that execute agent-directed tasks. More than 100 companies are participating, according to NVIDIA and launch partner SentinelOne, including Netskope and 1Password. NVIDIA'

Team Cymru researchers uncovered more than 80,000 LLM relay servers that can conceal the location and identity of users accessing frontier AI models.

Dragos has completed its acquisitions of asset-discovery company runZero and firmware security specialist NetRise, expanding its operational technology platform beyond threat detection and incident response. The deals closed alongside Accenture’s majority investment in Dragos. The companies initially announced transactions with a combined enterprise value of about $4.175 billion, subject to customary adjustments. Dragos said it will continue to operate independently, with fou

Rockwell Automation is testing advanced artificial intelligence to find and fix security weaknesses in the software and connected systems that underpin factories and other critical infrastructure. The industrial automation company has joined Anthropic’s Project Glasswing, an initiative that gives approved organizations controlled access to powerful Claude models for defensive cybersecurity research. Rockwell’s security teams will use Claude Mythos 5 to explore faster methods

Pindrop has launched BotStopper, a real-time detection system designed to help businesses distinguish AI voice agents and automated callers from humans. The technology arrives as autonomous agents increasingly contact banks, healthcare providers, insurers and retailers on behalf of consumers. The emerging challenge for contact centers is not simply blocking bots, but determining which automated callers are legitimate, who authorized them and what systems they should be permit

The autonomous cybersecurity system tests external exposure against internal architecture, security controls and existing risk data to help organizations separate reachable threats from scanner noise. Security teams rarely lack vulnerability data. The harder problem is determining which weaknesses attackers can actually exploit. Tuskira is targeting that gap with a new autonomous red teaming capability that evaluates an organization’s internet-facing attack surface and connec