AI Gatekeeper Launched To Secure Agentic AI

Operant AI Launches the real-time security application to boost Runtime Protection for Indian Businesses

Operant AI founders (L to R) Ashley Roof, Dr. Priyanka Tembey and Vrajesh Bhavsar

FinTech BizNews Service

Mumbai, April 17, 2025: Silicon Valley headquartered Operant AI, has launched AI Gatekeeper™, a real-time security application for live AI applications, agents and Agentic AI workflows—across Kubernetes, private, hybrid, and edge environments.

As organizations rapidly adopt autonomous AI agents and complex multi-agent workflows, especially in high-growth markets like India, security challenges have escalated. According to Deloitte’s State of GenAI report, over 80% of Indian organizations are exploring autonomous agents, with 50% focused on multi-agent setups that require minimal human oversight.

AI Gatekeeper™ goes beyond Operant’s existing 3D Defense capabilities, offering industry-first protections against rogue agents—such as trust scoring, agentic access controls, and threat blocking for Model Context Protocols (MCPs) and Non-Human Identities (NHIs).

Dependence on third-party vendors increases risks

Operant’s recent engagements with Indian enterprises and cybersecurity leaders highlight a strong interest in deploying AI agents but also significant reliance on third-party vendors for AI deployment, complicating data governance and security. Key concerns include data leakage, model poisoning, and rogue agent behavior. AI Gatekeeper directly addresses these issues, empowering enterprises to secure their agentic AI deployments at runtime across all platforms.

“The AI that we are now securing is a completely new beast compared to even two years ago,” said Vrajesh Bhavsar, Operant AI’s CEO and co-founder. He added that today RAG applications to AI Agents to AI Inference systems operate at a completely new scale, because of which AI can’t be secured in isolation. AI Gatekeeper can bring Operant’s unique defensive capabilities to everywhere customers are deploying AI, alongside critical new capabilities for protecting sensitive data and the rest of the application environment from the new attack surface that is being fueled by rapid Agentic AI adoption.”

Product Screenshot:

Gatekeeper creates space for faster AI innovation

“We are seeing three trends happening right now: First, incredibly fast deployment of AI models and AI Agents for novel use cases; second, adoption of new platforms beyond the traditional cloud providers; and lastly, the requirements and responsibilities for security, infrastructure, data infosec and AI converging,” said Raj Yavatkar, CTO of Juniper Networks. Operant has built a solution that helps teams protect their most business-critical transformations, while AI Gatekeeper makes it possible for AI-native teams to innovate securely at a completely new pace. 

AI Applications and agents are not only being built on cloud hyperscalers like Amazon EKS, Fargate, Bedrock, and similar services from Azure and Google Cloud, they are now expanding onto non-traditional platforms like Databricks, Snowflake, and Salesforce. The AI ecosystem – and the threats that come with it – are shifting closer to where the data that fuels AI actually lives. 

As a result, security and threat exposure are expanding while enterprises continue to add Agentic AI workflows that need to be controlled and secured by default to prevent new catastrophic failure modes. The adoption of new frameworks like MCP exposes enterprises to additional risks of breaches, like the tool poisoning vulnerability just reported last week, that require a fundamentally different security approach from traditional methods.

“Securing AI Agents is a critical priority for AI-native companies because you can’t hand off that level of autonomy at scale to these systems without appropriate controls in place,” said Martin Choluj, CISO of Clickhouse.

Operant’s AI Gatekeeper launch comes on the heels of Operant being named as a representative vendor in Gartner’s Market Guide for AI Trust, Risk, and Security Management (AI TRiSM), and mentioned in Gartner’s recent research note, “How to Secure Custom-Built AI Agents.”

Operant’s new AI Gatekeeper™ capabilities include:

• Runtime Defense for AI across Public, Private, and Hybrid Clouds
• Moving beyond Kubernetes, Operant’s 3D Runtime Protection is now available across public, private and hybrid cloud platforms
• Live comprehensive catalogs of all AI workloads, AI agents, tools, models, and all AI platforms that automatically update with the use of AI in an organization (includes providers such as OpenAI, Deepseek, Cohere, Anthropic, Hugging Face and more)
• Additional support for large data platforms, LLM and AI Agent platforms
• Defense analytics on all deployed AI defenses, providing unprecedented transparency into the threats that are being blocked at runtime 
• Cross-Platform Threat Modeling
• Cohesive AI Security Graphs mapping and flagging highest risk data flows between AI workloads, Agents, and AI APIs across all platforms on which they operate
• Out of the box mappings to OWASP Top 10 threat vectors for AI/LLMs and AI Agents, including sensitive data leakage, API key and secrets leakage, prompt injection, and data poisoning risks – with deep insights into actual threats and the workloads/APIs that are affected
• Advanced Threat Detection for AI Agents
• Supply chain risks for AI agents, with mapping of trust scores and boundaries
• Unauthenticated and unauthorized AI agent detection and defense with blocking capabilities
• Least privilege runtime execution and least permissioned trust boundaries for AI agents
• Protection for MCP and AI NHIs
• Beyond AI API security, AI Gatekeeper also includes supports for Model Context Protocol
• Detection and defense for both runtime and API access layers of agent tools built with MCP
• Expanded coverage for identity and access controls with support for AI NHIs