Aslan emerges from stealth with $20.8M to build AI undercover agents for national security

US startup Aslan publicly launched on September 1 after raising a combined $20.8 million across its seed and Series A rounds. The company is building so‑called “agentic AI” systems—autonomous software agents designed to operate as virtual operatives for national security missions.

Aslan’s agents sit on top of existing large language models but are wired into customized tools and workflows for defense and intelligence users. The idea is to help human analysts sift vast troves of open‑source data, spot subtle behavioral patterns and run simulated scenarios in virtual environments, effectively offering an AI “undercover assistant” rather than a physical robot.

According to the company, the new funding will accelerate R&D and expand pilot projects with US government customers. But the concept also raises familiar red flags: AI‑driven surveillance and information campaigns could easily cross ethical lines. Aslan says its systems are designed to operate only under human oversight and within legal constraints, reflecting a broader debate over how far governments should go in delegating sensitive decision‑making to autonomous AI agents.