Japanese AI startup Sakana AI has made a significant entry into the core of national defense and security. The company signed a research contract with the Defense Innovation Science and Technology Strategy Office of the Acquisition, Technology & Logistics Agency (ATLA). This multi-year project aims to fundamentally strengthen Command and Control (C2) systems by enabling AI to autonomously integrate and analyze the massive volumes of data generated across land, sea, and air domains.

The technological heart of this research lies in merging AI agents with Small Vision-Language Models (SVLM), a core strength of Sakana AI. AI agents are intelligent systems that can independently determine steps and execute tasks through trial and error to reach a specific goal. Meanwhile, SVLM is a lightweight version of AI capable of simultaneously processing images and text, designed to run on edge devices such as drones and handheld terminals. This allows drones to instantly analyze captured footage and report situations accurately on-site, even in environments with unstable communications.

In modern security, information power is a decisive factor between victory and defeat. However, humans have reached their limits in processing the flood of multimodal information—including visuals, audio, and sensor data—arriving from both cyber and physical spaces. By having AI instantly organize this information and propose optimal resource allocations, commanders can make swifter and more precise decisions. This initiative directly enhances the capabilities of C2 systems, which rely on rapid situational awareness.

Sakana AI has positioned Defense and Intelligence alongside finance as its most critical sectors, forming a dedicated domestic team to accelerate social implementation. This challenge—applying unique Japanese AI technology to bolster national security—is expected to serve as a model for open innovation, where startups support vital national infrastructure with cutting-edge technology.

Japanese AI startup Sakana AI has made a significant entry into the core of national defense and security. The company signed a research contract with the Defense Innovation Science and Technology Strategy Office of the Acquisition, Technology & Logistics Agency (ATLA). This multi-year project aims to fundamentally strengthen Command and Control (C2) systems by enabling AI to autonomously integrate and analyze the massive volumes of data generated across land, sea, and air domains.

The technological heart of this research lies in merging AI agents with Small Vision-Language Models (SVLM), a core strength of Sakana AI. AI agents are intelligent systems that can independently determine steps and execute tasks through trial and error to reach a specific goal. Meanwhile, SVLM is a lightweight version of AI capable of simultaneously processing images and text, designed to run on edge devices such as drones and handheld terminals. This allows drones to instantly analyze captured footage and report situations accurately on-site, even in environments with unstable communications.

In modern security, information power is a decisive factor between victory and defeat. However, humans have reached their limits in processing the flood of multimodal information—including visuals, audio, and sensor data—arriving from both cyber and physical spaces. By having AI instantly organize this information and propose optimal resource allocations, commanders can make swifter and more precise decisions. This initiative directly enhances the capabilities of C2 systems, which rely on rapid situational awareness.

Sakana AI has positioned Defense and Intelligence alongside finance as its most critical sectors, forming a dedicated domestic team to accelerate social implementation. This challenge—applying unique Japanese AI technology to bolster national security—is expected to serve as a model for open innovation, where startups support vital national infrastructure with cutting-edge technology.