AIShield, Alice, Enkrypt AI and Fiddler AI

About Alice

Alice, formerly known as ActiveFence, provides AI security, safety and trust products that test, protect and monitor generative AI applications, agents and foundation models across their lifecycle. Its WonderSuite platform includes pre-launch red teaming (WonderBuild), runtime guardrails (WonderFence) and production monitoring (WonderCheck), powered by its Rabbit Hole adversarial intelligence dataset.

About Enkrypt AI

Enkrypt AI sells an AI security platform with products for red teaming agents, enforcing runtime guardrails, turning policy into controls, auditing AI data, and scanning and gating MCP tool traffic. Enkrypt AI describes the platform as purpose-built to cover every threat, regulation, and modality.

Enkrypt AI Red Teaming finds failure modes across text, audio, and vision, including agents, tools, RAG, and MCP. Enkrypt AI Red Teaming turns findings into prioritized fixes and evidence-ready reports for security, risk, and compliance. Enkrypt AI Red Teaming adds closure evidence and regression tests to a CI gate. Findings can be promoted into Guardrails policies and monitored in production.^[3]

Enkrypt AI Guardrails is the runtime layer that approves, modifies, or blocks risky behavior across agents, tools, RAG, and MCP. Every enforcement decision is logged, and Enkrypt AI says each decision includes a policy id, policy version, and reason code. Enkrypt AI Guardrails takes SSO and IAM claims for role-aware and tenant-aware policies, including Okta, Azure AD, and custom JWT claims. Enkrypt AI Guardrails can apply guardrails inline in the MCP Gateway, not just in the model prompt. Enkrypt AI Guardrails is API-first and works with an existing model stack.^[4]

Enkrypt AI Policy Engine converts natural-language AI governance policy and regulatory text, including PDFs, into controls. Enkrypt AI Policy Engine enforces those controls across the platform with traceability that can be audited. Controls are reviewed and approved, then published to dev, stage, and prod environments. Policies are versioned, with diff review, staged rollouts, and rollback. Enkrypt AI Policy Engine combines internal governance requirements with external standards into a single control model. OWASP, NIST, and EU AI Act are included as defaults.^[5]

Enkrypt AI Data Risk Audit produces a sign-off packet for Security and Legal covering what data an AI system uses, what risk it carries, and what is approved for AI. Enkrypt AI Data Risk Audit evaluates the datasets a customer brings for a specific AI system and intended use. Intended uses include TRAIN, RAG, and EVAL. Enkrypt AI Data Risk Audit can block a dataset for TRAIN and approve it for RAG with conditions. Enkrypt AI Data Risk Audit works across text, audio and images, in multiple languages. Enkrypt AI Data Risk Audit connects to Jira and ServiceNow for remediation and approvals, and is re-run when datasets or connectors change.^[6]

Enkrypt AI MCP Scanner discovers tools, analyzes capabilities and permissions, and surfaces high-risk exposure. MCP Scanner covers configuration, permissions, tool schemas, endpoints, and common vulnerability patterns on MCP servers and tool surfaces. MCP Scanner scans internal MCP servers during development and before rollout, not just public MCPs. MCP Scanner scores severity on exploitability and impact, and on exposure. MCP Gateway is open source and sits inline between agents and MCP servers to approve, modify, or block tool calls. Enkrypt AI adds enterprise capabilities to MCP Gateway such as centralized policy management, tenant-aware enforcement, reporting, and support. MCP Gateway runs as a sidecar or gateway in the tool path, in Kubernetes, edge, or service mesh patterns. MCP Scanner findings inform an approved registry and policy defaults, and MCP Gateway enforces those decisions at runtime.^[8]

Enkrypt AI says its customer-facing agents offering targets teams shipping customer-facing chat or voice agents in regulated industries such as Finance, Healthcare, Insurance, and Government. Enkrypt AI produces an Agent Release Packet, a single, forwardable document that gives every stakeholder the evidence they need to approve an agent. Enkrypt AI says teams that used to spend 4 to 6 months in review loops ship in 30 days. Enkrypt AI describes a large insurance services provider that used Enkrypt AI to secure its customer-facing AI-powered Chatbot app. Enkrypt AI describes a financial organization that scaled its tax preparation services through an AI-enabled tax chatbot. Enkrypt AI describes a medical device company that wanted to deploy an AI app for personalized medicine adherence. NetApp partnered with Enkrypt AI to ensure data readiness for Gen AI and to remove data risks on its platform.^[9][10]

Enkrypt AI lists SOC 2 Type II and ISO 27001 among its security and compliance credentials. Its address is 1660 Soldiers Field Rd. Brighton, MA 02135.^[11]

About Fiddler AI

Fiddler AI sells an AI observability and security platform that monitors machine learning models and AI agents, enforces policy guardrails, and supports AI governance and compliance. Fiddler AI was founded in 2018. Fiddler AI raised $30 million in Series C funding led by RPS Ventures, bringing its total funding to $100 million. CrowdStrike announced plans to acquire Fiddler AI to deliver agent observability for AI trust.^[2]