DARPA Selected Xint for Military AI Security
The agency tapped the startup to automate application security for messaging platforms using advanced AI models.
Updated on Sept. 29, 2026 in Cybersecurity

Live Poll
Do you trust autonomous AI systems to secure sensitive government or personal communications?
In September 2026, DARPA selected the startup Xint to lead research into autonomous AI application security for military messaging. The project stems from Xint winning a portion of a $29.5 million prize pool during the two-year Artificial Intelligence Cyber Challenge.
Why it matters
DARPA aims to fortify Department of War messaging systems against potential foreign and external eavesdropping. The partnership leverages Xint's ability to analyze software code and binaries to ensure secure communication channels.
Xint employs frontier LLM models to reverse engineer binaries and examine system components between applications and the kernel. The company operates as a SaaS solution and is currently developing a version of its service for internal data centers.
The players
DARPA
Established in 1958, this agency oversees the development of emerging technologies for military use by the United States.
Xint
A spinoff from the company Theori, this firm provides a SaaS solution for software supply chain risk assessment.
Theori
This security firm served as the parent organization from which the Xint startup emerged.
The details
Originating from the company Theori, Xint specializes in software supply chain risk assessment. Its technology provides deep analysis of source code and compiled binaries to identify security flaws before they can be exploited.
Timeline
DARPA was established in 1958.
Xint launched a software supply chain service in September 2026.
The Tech Race
This move reflects a broader industry shift toward integrating LLM-based autonomous security into critical software supply chains. It replaces manual code auditing with scalable, automated analysis to keep pace with modern cyber threats.
While the technology is currently focused on military applications, Xint's shift toward data-center-based deployment may eventually improve security standards for private enterprise software. Users in government sectors will likely see more automated, robust protection for the communication platforms they access daily.
The takeaway
Automating application security is becoming essential as AI models are increasingly used to both secure and threaten software infrastructure. Organizations looking to harden their systems should prioritize supply chain transparency and the analysis of compiled binaries.
Further reading
For broader trends in digital defense and federal initiatives, visit the Cybersecurity section.
Live Poll
Do you trust autonomous AI systems to secure sensitive government or personal communications?










