The U.S. Army announced on October 23 the selection of two new partners for the so-called “Project Griffin.” These are Amazon Web Services (AWS) and World Wide Technology (WWT). As part of a pilot project with a total value of 2.6 million U.S. dollars, these companies are to help advance AI-enabled cyber defense. With this initiative, the U.S. Army aims to establish autonomous operational capability that allows soldiers to defend their own networks more effectively.
This is intended to eliminate typical vulnerabilities that arise from manual data analysis and human intervention. Specifically, the Army is seeking an autonomous, action-ready AI platform that enables cyber defenders to neutralize threats much faster. Project Griffin is one of three accelerated initiatives under ARDS (Army Rapid Defense of Cyber Systems)—an Army initiative that drives the deployment of AI-enabled cyber defense capabilities.
Project Griffin and Project ARDS
ARDS is led by the Army’s Chief Cyber Advisor in close collaboration with the Deputy Secretary of the Army for Acquisition, Logistics and Technology, Army Cyber Command, the Chief Information Officer (CIO), and other Army leaders. Together they identify, procure, and implement promising technologies for the Army’s cyber defense. The operational leadership of Project Griffin lies with the Project Manager for Cyber Warfare (under the Portfolio Management for Reconnaissance and Spectrum Warfare). The contractual handling is carried out by the Army Contracting Command at Aberdeen Proving Ground, Maryland.
“Collaboration with the private sector is a matter of national security for the Army, as the private sector is pushing rapid innovations in the areas of superintelligence, cloud architecture, and capabilities for cyber operations,” explained Colonel David Bergmann, Project Lead for Cyber Warfare. “Thanks to our flexible procurement procedures, we could swiftly pursue solutions in the realm of superintelligence that strengthen our cyber forces. Cooperation with industry ensures that the Army maintains its technological edge and secures modern digital dominance over potential adversaries.”
Greater Availability Without Additional Risks
For practical deployment, it will be crucial how reliably the selected solutions operate under realistic conditions. Automated detection of suspicious activities alone is not enough: the systems must also be able to distinguish between actual attacks, technical malfunctions, and unusual but legitimate user behavior. Faulty countermeasures could otherwise disrupt important connections or hinder access to needed information. Therefore, clearly defined intervention rights, traceable decisions, and the possibility of human oversight are essential prerequisites for a rollout.
The pilot project provides the opportunity to verify these requirements before broader use. In addition to detection speed, false alarm rates, integration into existing networks, and the operational burden should be considered. Equally important is protecting the deployed AI from manipulated input data and targeted deception attempts. Whether the tested approaches provide an operational added value will be measured by whether they effectively relieve cyber forces and at the same time increase the availability of military networks, without creating additional, hard-to-control risks in operation.
Author: André Forkert, Military Affairs Editor
