Alterion launches Aquila to extend AI policy control to employee devices
Alterion has launched Aquila, a product it says lets enterprises control AI activity on employee devices and extend policy enforcement beyond cloud systems.
Alterion said on Aug. 18, 2026, that it launched Aquila, a product designed to help enterprises control and govern AI use on employee devices.
The company said Aquila is intended to address what it described as a blind spot in enterprise AI governance: work done on laptops and workstations outside cloud-based controls. Alterion said developers and employees are using coding assistants, browser-based AI tools and local AI applications on endpoints, and that some higher-risk AI actions now happen on those devices.
Alterion said the product is meant to give security, risk, compliance and engineering teams a way to apply policy where AI actions begin. The company said the system can inspect AI-driven actions and data before they leave the device, which it said gives organizations a more unified way to enforce AI policy across environments.
According to the release, Aquila works at three layers: an allowlist for approved tools, blocking sensitive data from leaving in an AI request, and discovery of AI activity outside sanctioned tools. Alterion said the product extends its Draco runtime control plane from cloud-based AI systems to endpoints.
The release said Alterion describes itself as a runtime control platform for enterprise AI and said Aquila is available today for enterprise customers. It also said the company’s platform includes Draco, Helix and Aquila as parts of its runtime control plane for what it calls the agentic enterprise.
Alterion included two short comments from company representatives in the release. One said sensitive AI work is happening on employee devices. Another said Aquila is designed to decide whether information leaves the device at all.
The company said users can learn more through its blog, request a demo and review technical details on the Aquila product site. No independent corroboration was included in the provided material.
Sources
Editorial note: This article was prepared with AI assistance and includes source attribution for reader review.
Comments ()