Army Cyber Command Developing Autonomous AI Agents for Offensive Cyber Operations Under ECHOTRIBBLE Program
Documents obtained by The Intercept show Army Cyber Command is developing ECHOTRIBBLE, a platform that would turn AI agents into autonomous tools to navigate adversary networks and "neutralize" targets, including
ARCYBER is shifting from human-operated intrusion toward self-directed agents without a settled answer on containing collateral damage. Requirements for "minimal human oversight," self-deletion, and air-gapped reach mean operators would have little ability to recall or audit an agent once launched, and replication across iPad, Apple desktop, and Windows 11 targets could carry it through civilian networks that adversaries use as staging points. Contractor whitepapers and any ARCYBER statement on safeguards are the next indicators. Confidence is moderate: two document-based accounts agree on intent, but both trace to The Intercept's document review, only the SAM.gov notice independently confirms the program exists, ARCYBER has not commented, and no working tool has been reported. The program may instead be an early market-sounding exercise whose autonomy is aspirational, with human approval gates added before fielding.
4 sources
- The Army Wants to Sic AI Agents on Its Enemies. What Could Go Wrong? -
The Intercept - Army Cyber Command's ECHOTRIBBLE would turn AI agents into self-spreading cyberweapons -
The Decoder - US Army is developing AI agents that could strike enemy networks and vanish without a trace -
Help Net Security - Agentic C2 Framework Tool Assessment Event -
TechBids