Leaked documents reveal US Army's ECHOTRIBBLE autonomous cyber warfare project
Leaked military and vendor documents describe ECHOTRIBBLE, a nascent U.S. Army Cyber Command initiative designed to deploy agentic AI to autonomously infiltrate adversary networks and neutralize digital targets.
Leaked military and vendor documents have revealed details regarding ECHOTRIBBLE, an initiative managed by U.S. Army Cyber Command alongside the Cyber Fusion Innovation Center and the Army Cyber Technology and Innovation Center, according to reporting by The Intercept's Sam Biddle and further reporting from Gizmodo. Officially designated the Agentic C2 Framework Tool, the program focuses on building self-directed AI software capable of infiltrating networks, locating digital targets, and neutralizing them with minimal human intervention.[1][3][4][5]
The software is designed to operate in both cloud environments and air-gapped systems disconnected from the public internet, behaving much like an autonomous computer worm. The AI agents are built to silently hop across devices on a foreign network, carry out offensive actions such as disabling systems or deleting target data, and subsequently self-destruct to minimize detection. Vendor materials clarify that offensive strikes are the primary focus of the project.[2][4]
The disclosures have prompted serious concerns among cybersecurity experts and researchers regarding the risks of giving autonomous systems mandates to neutralize targets. Critics warn that the unpredictable nature of agentic AI models could lead self-propagating tools to accidentally cross into civilian infrastructure, cause collateral damage, or ignite rapid cyber escalations without meaningful human oversight.[4]
Key facts
- Leaked documents describe ECHOTRIBBLE, an autonomous cyber warfare initiative managed by U.S. Army Cyber Command, the Cyber Fusion Innovation Center, and the Army Cyber Technology and Innovation Center.
- The program is officially designated the Agentic C2 Framework Tool and primarily focuses on offensive strikes.
- The software is built to operate like an autonomous worm, hopping across foreign devices in cloud or air-gapped systems, deleting data or disabling systems, and then self-destructing.
- Cybersecurity researchers warned that unpredictable AI agent behavior could spread to civilian networks and escalate conflicts without human intervention.
Sources · 5 sources
- TE
TechmemeArticle ·
Leaked documents detail ECHOTRIBBLE, a nascent US Army Cyber Command program for agentic AI to autonomously spread across networks to find and eliminate targets (Sam Biddle/The Intercept) Sam Biddle / The Intercept : Leaked documents detail ECHOTRIBBLE, a nascent US Army Cyber Command program for agentic AI to autonomously spread across networks to find and eliminate targets — Documents obtained by The Intercept show the military is trying to deploy autonomous AI agents to hack its adversaries.
Open source - TH
Tom's Hardware@tomshardwarePost on X ·
U.S. Army developing program that would deploy AI agents to autonomously seek and destroy targets in enemy digital infrastructures https://t.co/2RmawccqC9
Open source - TE
Techmeme@TechmemePost on X ·
Leaked documents detail ECHOTRIBBLE, a nascent US Army Cyber Command program for agentic AI to autonomously spread across networks to find and eliminate targets (@samfbiddle / The Intercept) (Visit Techmeme dot com for the link and full context!)
Open source - WE
webulite@webulitePost on X ·
Leaked military documents and vendor materials revealed details surrounding ECHOTRIBBLE, an autonomous cyber warfare initiative managed by U.S. Army Cyber Command, the Cyber Fusion Innovation Center, and the Army Cyber Technology and Innovation Center. Officially designated the Agentic C2 Framework Tool, the project focuses on developing self-directed AI software capable of navigating digital networks, locating targets, and neutralizing them with minimal human intervention. Designed to operate in cloud environments and air-gapped systems disconnected from the public internet, the software functions similarly to an autonomous computer worm. The AI agents are structured to silently hop across devices on a foreign network, carry out offensive actions such as deleting target data or disabling systems, and then self-destruct to minimize detection. Internal vendor documents clarify that offensive strikes are the primary focus of the program. Cybersecurity experts and researchers expressed serious concern over the initiative, pointing out the risks of deploying autonomous systems with mandates to neutralize targets. Critics highlighted that agentic AI models frequently exhibit unpredictable behavior, raising fears that self-replicating tools could accidentally traverse civilian infrastructure, cause unintended operational collateral damage, or trigger escalating cyber conflicts without meaningful human oversight.
Open source - GI
Gizmodo@GizmodoPost on X ·
Documents Reveal Military Plan for Agentic AI That Hacks U.S. Adversaries https://t.co/PXrPZ4Z4Pl
Open source