AppViewX Enhances Agent Identity Security to Combat Shadow AI Risks
AppViewX has upgraded its Agent Identity Security solution to provide enterprises with comprehensive discovery, governance, and runtime enforcement for all AI agents, including unsanctioned 'shadow AI'.
AppViewX has significantly bolstered its Agent Identity Security platform, introducing new capabilities designed to tackle the escalating risks associated with the proliferation of artificial intelligence agents within enterprise environments. The enhanced solution aims to provide Chief Information Security Officers (CISOs) with the tools to discover every AI agent, whether officially sanctioned or operating in the shadows, and to enforce strict runtime controls over their actions. This update also introduces quantum-resilient agent identities, ensuring that the cryptographic underpinnings of enterprise trust remain secure as the technology landscape evolves.
The rapid expansion of AI agents presents a novel identity and access management challenge. Each agent can possess its own models, credentials, privileges, and connections to critical enterprise systems, operating autonomously or on behalf of users at a speed that traditional access control mechanisms are ill-equipped to handle. The need to secure the credentials and cryptography that authenticate these identities is paramount, especially as organizations prepare for a future where quantum computing could undermine current encryption standards.
AppViewX positions its approach to agent security as a fundamental identity and access problem, requiring a departure from retrofitting existing human identity architectures. The Agent Identity Security platform is purpose-built from the ground up at the identity layer, offering what the company describes as the industry's most comprehensive discovery, governance, and runtime security for a wide array of agent types, including coding agents, productivity agents, endpoint agents, SaaS agents, and custom-built solutions.
Key enhancements include expanded discovery capabilities for an 'AI Bill of Materials' (AIBOM), which now identifies browser-based and short-lived, script-based shadow agents. This AIBOM extends to the skills agents can access, allowing security teams to assess potential risks and flag unsafe capabilities. AppViewX combines its own lightweight endpoint Guardian Agent with API integrations across EDR, SaaS, and cloud platforms to ensure comprehensive detection, aiming to leave no shadow agent undetected.
Furthermore, the platform introduces a new 'MCP Gateway' for governing both sanctioned and shadow Machine Control Protocol (MCP) servers, continuously assessing their risk posture. It implements just-in-time access to eliminate standing privileges, granting agents access only when and for the duration needed, based on context. Data redaction is also supported through built-in controls or integrations with existing data security tools.
A critical new feature is the 'Agent Kill Switch,' designed to terminate an agent and its active sessions comprehensively. This can be triggered automatically by detected events, through runtime policies, or on demand. The kill switch integrates with the Shared Signals Framework (SSF) for broader threat intelligence correlation and can act on signals from third-party security tools. Unlike other solutions, it covers both sanctioned and shadow agents and does not rely on an MCP gateway.
AppViewX has also enhanced cost and compliance controls for AI programs, providing visibility into AI token usage, trends, and costs, with policy thresholds to prevent runaway spending. To aid in regulatory compliance, the platform enables organizations to assess and demonstrate alignment with various standards, including OWASP Top 10 for Agentic Applications, MITRE ATLAS, GDPR, HIPAA, ISO 27001/42001, NIST SP 800-53 Rev. 5, NIST AI RMF, the EU AI Act, SOC 2, and SEC Cyber Disclosure.
Finally, AppViewX leverages its core Public Key Infrastructure (PKI) and Certificate Lifecycle Management (CLM) expertise to issue quantum-resilient agent identities. These identities provide an additional layer of trust when authenticating via an Enterprise Identity Provider (IdP), chaining to the customer's AppViewX-managed PKI to offer a single trusted root, a tamper-evident audit trail, and a centralized method for building quantum-resilient identities.
The latest enhancements to AppViewX's Agent Identity Security solution introduce quantum-resilient agent identities, aiming to protect enterprises against future cryptographic threats. This update also expands the AI Bill of Materials (AIBOM) to include agent skills, providing a more comprehensive view of potential risks and unsafe capabilities.