Cynerio

Artificial intelligence is a layer on the platform the product now runs in. Cynerio's medical device security ships inside Axonius for Healthcare, part of the Axonius asset intelligence platform, which discovers devices by passive deep packet inspection at a core switch, classifies them "with over 150 attributes per asset", reads medical protocols such as HL7 and DICOM, and correlates device behavior in a threat detection engine; none of those functions is described as learned.

The AI Axonius describes is platform-wide: a natural language engine for asking questions, an "AI-driven risk engine" for prioritization, and an Axonius AI Agent built on Amazon Bedrock. Discovery, inventory, CMMS and ticketing integration and enforcement playbooks carry value without it.

The product claims complete discovery, and automated enforcement runs with no approval step described. The Axonius for Healthcare page says it will "Discover and classify every medical device with over 150 attributes per asset" and promises "Complete visibility"; Axonius's October 2025 launch release says it will "Automatically discover and classify every connected medical, IoT, and OT device".

A buyer who takes that at its word has no reason to reconcile the inventory against the CMMS, which is the step that would catch a device the platform missed or misclassified. The page also offers to "Automate alerting, ticketing, and enforcement using pre-built or custom incident response playbooks", and the launch release lists isolating a compromised device from the network among response actions. Nothing on the product page, the healthcare documentation or the launch materials describes who approves an enforcement action against a device in clinical use, or what happens after a wrong one.

The approach and a supplier are named; models and versions are not. Axonius for Healthcare's documentation describes discovery through the Axonius Network Inspector, a server appliance on a core switch SPAN port that analyzes mirrored traffic by deep packet inspection, with medical protocol support including HL7, DICOM and RTLS.

Axonius says its AI Agent is grounded in each customer's own asset data and runs isolated per customer on Amazon Bedrock, built with Amazon Bedrock AgentCore (release of 29 September 2026). The model behind the agent, the method of the "AI-driven risk engine", and how device classification works are not described, and no model version or change record is published. Axonius publishes product release notes.

The host and the model layer's supplier are named. Axonius's security program says it hosts its service in Amazon Web Services "in a single-tenant manner", and its 29 September 2026 release says each customer's Axonius AI Agent is kept isolated using Amazon Bedrock. The Axonius Trust Center lists a subprocessor document. The retention and training terms that apply to data sent to Bedrock are not published, and the model the agent uses is not named. The Network Inspector appliance performs deep packet inspection inside the hospital, so what it sends to the Axonius cloud determines what clinical network data leaves the site.

No named healthcare deployment and no performance claim a reader can check. Neither the Axonius for Healthcare page, its documentation, the October 2025 launch materials nor the July 2025 acquisition release names a hospital using the product or gives a detection, classification or outcome figure. Axonius reports being named a Strong Performer in the Forrester Wave for Proactive Security Platforms, Q3 2026, an analyst assessment of the wider platform. Cynerio's earlier Ponemon research with 517 US health systems sized the problem of connected device risk rather than measuring the product.

General security assurances, without the answers AI raises. Axonius says each customer has "a dedicated, isolated environment", that data is encrypted in transit and at rest, and that each customer's AI Agent is isolated on Amazon Bedrock.

The healthcare product inspects packets on the clinical network, which carries patient data between devices and systems, and nothing says what the Network Inspector keeps, what reaches the cloud, how long it is retained, whether any of it trains models, or how it is de-identified. Nothing covers guardrails on the AI Agent or how an AI error is handled.

HIPAA controls are attested; business associate status is not stated. Axonius's security program cites a Type 1 attestation for the HIPAA Security Rule and HITECH breach notification, and the Axonius Trust Center lists HIPAA among its frameworks.

Neither page, nor the Axonius for Healthcare page, says whether Axonius acts as a business associate for the healthcare product, offers an agreement, or at which tier, although the product inspects clinical network traffic that can carry protected health information.

Recognized audits are named in the company's own material, with documents behind the trust center's gates. Axonius's security program cites a SOC 2 Type 2 attestation from Schellman, ISO 27001 and a HIPAA Security Rule attestation, and says third party penetration testing is commissioned regularly.

The Axonius Trust Center lists ISO/IEC 27001:2022 with its statement of applicability, SOC 2, CSA STAR, FedRAMP Moderate, TX-RAMP Level 2, IRAP and C5, sorting documents into public, private and request-only without showing the report dates on its open face.

No device claim; the product works the device regime for the customer's fleet without stating its own position. Axonius for Healthcare is security software with no FDA clearance and none claimed. Its page offers "fleet-level risk scores, MDS2 and FDA data, and device utilization trends" and native support for MDS2 among medical protocols and data. No page states an intended purpose or says how enforcement against a device in clinical use fits the hospital's device risk management.

Nothing is published on how the AI or detection models are governed or tested. Axonius describes its AI Agent, natural language engine and AI-driven risk engine, and the healthcare threat detection engine that correlates device behavior, but no governance framework, evaluation, error handling or review process appears on the product page, the healthcare documentation, the trust center or the security program page.

The failure that matters is a device misclassified or a legitimate behavior read as a threat, followed by an automated enforcement action on a device in clinical use; no result is published by device type or manufacturer.

Nothing on the product page, the healthcare documentation, the trust center or the security program page describes what happens when a device classification, a risk score or an automated enforcement action is wrong: no accuracy or false positive figure, no warranty or remediation commitment, and no route to correct an output. Enforcement can run from playbooks against devices in clinical use, which is where an error has consequences for patient care.

It has no clinical record surface of its own; it connects to the clinical engineering and security systems a hospital runs. Axonius for Healthcare names CMMS integrations with Medimizer, e-Quip and Medusa, and Axonius's documentation covers a TRIMEDX adapter; the product page says it unifies CMMS, vendor catalogs, EHR and RTLS data with network telemetry, and automates tickets and enforcement through network access control and firewalls. The CMMS connections are described as reads into the Axonius inventory; no two-way synchronization, named record system or production site is shown.

Options, residency and isolation are stated; the processing path is not. Axonius hosts its service in Amazon Web Services in a single-tenant environment for each customer, lets customers choose the AWS region for their instance, and lets customers host their own instance instead. Axonius for Healthcare adds a Network Inspector appliance inside the hospital that inspects mirrored traffic at a core switch. Nothing says which processing leaves the chosen region, such as AI Agent calls through Amazon Bedrock or support access.

Pricing from Axonius waits for a sales conversation. The Axonius for Healthcare page, its documentation and the launch materials publish no price, tier, unit of charge or route to start. A reseller's AWS Marketplace listing for the Axonius Asset Cloud, sold by Carahsoft by private offer, prices the core platform at $29,745 a year for 500 assets; that is the reseller's listing for the wider platform, not an Axonius price for the healthcare product.

The setting is stated clearly, with nothing named to validate it. Axonius for Healthcare is "Purpose-built for hospitals" for security and biomedical teams, covering medical, IoT, OT and IT assets in one inventory with medical protocol support. The product page, documentation, launch release and acquisition release name no hospital using it, and Cynerio's former customer references are not carried onto the Axonius estate.