MetaGuard AI launches Vanguard for agentic cyber defense

MetaGuard AI Launches Vanguard to Harden Cyber Defenses Through Agentic AI-Driven Attack-and-Defense Simulations
🕧 5 min

Quantum Ventura spin-off brings affordable, automated cybersecurity to small and mid-sized businesses in the U.S. and Japan, from US$4,999 a year; LK Partners named distributor for Japan and the Asia-Pacific region.

MetaGuard AI, a spin-off of defense technology developer Quantum Ventura Inc., today launched two AI security products: CNRT Vanguard, which stages simulated cyberattacks in isolated test environments to assess defensive responses, and CNRT CodeLens, which checks software code for security flaws and offers fixes. Available immediately, both products expand CyberNeuro-RT (CNRT) into a comprehensive cybersecurity platform.

Michigan State University, IIT Palakkad, OpenPOWER Foundation (a Linux Foundation member), SRM University-India, and Objection Automation Inc. are running proof-of-concept trials. In a controlled test reported by Quantum Ventura, Vanguard’s AI defender agents shut down a password-guessing and network-scanning attempt in under a minute — closing the targeted port before the attacker agents could strike again.

Read more – Ebix partners with intentyx for AI agent security

Vanguard: rehearse attacks and identify defensive gaps

Vanguard (from US$4,999 a year) runs automated exercises in which AI agents act as attackers and defenders. Using a prebuilt lab or a replica of the customer’s environment rather than live systems, the agents attack, detect, and respond in a continuous cycle. Exercises run in isolated, disposable virtual machines with safeguards. Vanguard scores each exercise and identifies gaps in detection.

Vanguard records attack and defense actions as labeled training data generated by the simulations. Customers can export it in CNRT-compatible or custom formats and use it to retrain CyberNeuro-RT’s machine-learning models or other detection models.

Read more –Penguin Solutions launches high-bandwidth MRDIMMs for AI

CodeLens: catch security flaws before software release

CodeLens (from US$3,999 a year) uses AI agents to analyze software source code in an isolated virtual machine where AI reviewer agents challenge each finding to reduce false alarms. CodeLens then prioritizes the vulnerabilities based on how easily attackers could exploit them. Fixes are offered for manual review or automatic application.  Both products support Claude, GPT, or fully local AI models.

“Vanguard’s closed-loop red-team/blue-team architecture strengthens customers’ defenses with data from realistic attacks, so their defenses can be hardened with every exercise,” said Srini Vasan, President and CEO of Quantum Ventura.  “Vanguard and CodeLens add more firepower to the unified CNRT ecosystem at affordable price points.”

Read more – Clarivate brings trademark data to Microsoft Copilot

Japan and Asia-Pacific distribution

LK Partners, Tokyo, has been appointed master distributor for Japan and Asia-Pacific.

“As cyberattacks increase, CNRT gives Japan’s small and medium-sized businesses affordable access to enterprise-grade AI protection,” said Yasuyuki Morikawa, CEO of LK Partners.

Write to us editorial@demandmediaagency.com. to learn more about our exclusive editorial packages and programmes.

  • PR Newswire

    What began as a wire service in 1954 has evolved into one of the largest global distribution networks. PR Newswire, now part of Cision, gives communicators direct access to journalists, editors, and digital outlets, helping stories break beyond borders and shape conversations in real time.

Recommended Reads :