SAN FRANCISCO, March 24, 2026 /PRNewswire/ - AccuKnox today announced AI-Security 2.0, a comprehensive framework for Identity-powered, Zero Trust platform to secure AI models, Agents, and Data.
What's New in AI-Security 2.0?
AccuKnox AI-Security 2.0 includes eight integrated capabilities — six generally available at launch, three in active beta:
- AI Security Posture Management (AI-SPM)
- AI Detection & Response (AI-DR)
- AI Red Teaming
- Prompt Firewall
- AI Pen Testing / CTEM
- AI-Data Security Posture Management (AI-DSPM)
- AI-GRC (Beta)
- AI Identity Controls (Beta)
- Beta (Beta)
AccuKnox AI-SPM continuously discovers and maps every model endpoint, notebook, MLOps pipeline, and agent toolchain across teams and clouds. Security teams get a real-time inventory of what is running, where it runs, and who owns it. This allows organizations to detect shadow AI usage and enforce governance early.
AI-DR - Detect What Matters, Fast
AccuKnox AI-DR analyzes CloudTrail, Azure logs, and runtime telemetry to detect risky AI infrastructure activity such as unexpected resource creation, new region deployments, public exposure, privilege escalation, and abnormal GPU usage. Automated response workflows allow teams to remediate risks immediately.
AI Red Teaming - Break It Before Attackers Do
AccuKnox runs automated adversarial testing against hallucinations, prompt injection, toxicity, and unsafe code generation. Security teams receive repeated test results that help measure model behavior and risk before release.
Prompt Firewall - Stop Injection at the Gate
AccuKnox Prompt Firewall inspects every prompt and model output using policy-based controls. It blocks malicious instructions and enforces least-privilege access at the AI gateway without requiring changes in the application code.
Unique Differentiation
- Upstream caller sequence aware authorization for multi-agent workflows
- Runtime sandboxing with least-privilege tool execution
- Unified visibility across hybrid and multi-cloud environments
- Open standards using SPIFFE and OpenFGA
- Available on AWS, Azure, Red Hat, and Oracle Cloud Marketplaces
Deployment
AccuKnox deploys the control plane across AWS, Azure, GCP, on-premises Kubernetes, VMs, containers, and serverless environments with native integrations for MCP servers, AI gateways, and leading LLM providers including Amazon Bedrock, Gemini, Ollama, and vLLM. Certified by NVIDIA Inception and AWS Partner Network.
What Our Customers, Partners, Analysts Are Saying
Clients
"AI is quickly becoming the core infrastructure for modern enterprises, but the security model around it is still catching up. Models, agents, and data pipelines introduce new identity and runtime risks that legacy cloud security tools can't address. AccuKnox AI-Security 2.0 takes a meaningful step forward by applying Zero Trust principles directly to the AI layer - bringing visibility, runtime protection, and identity-driven governance to the environments where AI actually runs. That kind of architectural approach will be essential as organizations move AI from the lab into production."
Golan Ben-Oni, CIO and CISO, IDT Corporation
"Our mission at Tible Group is to keep IT human, ensuring that innovation and security grow together. By partnering with AccuKnox as early adopters, we've seen firsthand how their AI-Security 2.0 platform provides the runtime visibility and controls needed to manage rapidly evolving AI workloads. Built on trusted CNCF projects like KubeArmor, it allows our clients to innovate fearlessly while keeping emerging risks in check."
Merijn Boom, Founder, Tible Group & SecDesk
"AI is rewriting the rules of business — and with that power comes serious responsibility. AccuKnox AI-Security 2.0 is built for organizations that refuse to compromise on security, compliance, or governance. Rooted in deep AI and Kubernetes expertise and powered by leading CNCF open-source projects like KubeArmor and ModelArmor, it delivers agentic AI security, robust identity controls, and deployment flexibility — so you can innovate fearlessly while keeping emerging risks in check."
Bala Kannan, Founder and CEO, Workb.ai
Analysts
"The enterprise AI stack is maturing faster than its security model. Models, agents, and data pipelines create a distinct risk surface — one involving identity, runtime behavior, and data exposure — that legacy cloud security tools weren't architected to address. AccuKnox AI-Security 2.0 takes a principled approach: applying Zero Trust at the AI layer to provide runtime protection, visibility, and identity governance where the exposure actually exists. Organizations serious about moving AI into production will need exactly this kind of infrastructure-level thinking."
Dr. Edward Amoroso, Chief Executive Officer, TAG Infosphere, Inc., Former SVP/CSO, AT&T (ret.)
"With AI emerging as one of the most consequential drivers of enterprise transformation, security, governance, and compliance are mission-critical. AccuKnox AI-Security 2.0 stands out as a highly advanced platform, building on deep expertise in Kubernetes runtime protection and widely adopted CNCF open-source projects such as KubeArmor and ModelArmor."
Janakiram MSV, Principal Analyst, Janakiram & Associates
"AI is rapidly becoming one of the most consequential technologies driving enterprise transformation. As organizations scale AI adoption, ensuring robust security, governance, and compliance becomes mission-critical.
Lawrence Pingree, SACR, Head of Data and AI Security Research
Partners
"AccuKnox Team deserves huge congratulations on launching AI Security 2.0!. Their innovative approach to addressing security, governance, and compliance challenges is game-changing for organizations embracing AI.
Mohan Kumar TL, Director, Netpoleon
"At Mainsail, protecting customer workloads is foundational. As organizations deploy AI at scale, the next frontier of security is protecting AI models, training data, and user interactions themselves. Through our partnership with AccuKnox AI-Security 2.0, we are extending Starlight's security posture beyond infrastructure into the AI layer, helping organizations safely deploy and operate AI workloads in even the most sensitive environments."
Brad Sollar, CTO, Mainsail Industries
"Given that AI is one of the most potent business transformation tools, it is imperative that organizations leverage it to ensure safety, security, compliance and governance. AccuKnox AI-Security 2.0 is clearly the most powerful platform
offerings in the industry. AccuKnox is an ideal partner for organizations in their journey to leverage the power and potential of AI"
Avinash Advani, Founder & CEO, CyberKnight
AI is reshaping how organizations innovate and compete. However, realizing its full potential requires strong safeguards around security, governance, and compliance. AccuKnox AI-Security 2.0 exemplifies a new generation of platforms designed to meet these challenges.
Paul Calatayud, CSO, Archipelo, previously CISO, Aqua Security; CSO, Palo Alto Networks
See AccuKnox AI Security 2.0 Live at RSAC 2026
AccuKnox is demonstrating AI-Security 2.0 live at RSA 2026, North Expo Booth #N-5154, Moscone Center. The demo covers multi-cloud agentic authorization enforcement, real-time prompt firewall in action, and full AI asset discovery across a heterogeneous environment.
Book Private Briefing with AccuKnox at RSA #N-5154
About AccuKnox
AccuKnox delivers a Zero Trust Security platform for AI, API, Application, Cloud, and Supply Chain Security. Incubated out of R&D innovator, SRI International (Stanford Research Institute), AccuKnox holds seminal Zero Trust security patents and is backed by top-tier investors including National Grid Partners, Dolby Family Ventures, Dreamit Ventures, Avanta Ventures, and 5G Open Innovation Lab.
Learn more at https://accuknox.com/
Media Contact
Syed Hadi, AccuKnox, 91 8015466708, [email protected], https://accuknox.com/
SOURCE AccuKnox
Share this article