Inside Anduril and Meta’s quest to make smart glasses for warfare

The Future of War is Worn: How Anduril and Meta Are Building the Ultimate Soldier’s Smart Glasses

In the not-so-distant future, a soldier crouched in a dusty Afghan valley may lock eyes on an enemy vehicle—and with a glance and a whisper, summon a swarm of drones to neutralize the threat. This isn’t science fiction. It’s the vision driving a bold new collaboration between defense-tech startup Anduril and tech giant Meta, as they race to develop augmented-reality (AR) smart glasses that could redefine modern warfare. These aren’t just gadgets—they’re the next evolution of the soldier, merging human intuition with artificial intelligence in a seamless, wearable command center.

At the heart of this transformation is Quay Barnett, a former Army Special Operations commander now serving as vice president at Anduril. His mission? To optimize “the human as a weapons system.” That phrase, once the stuff of cyberpunk novels, is now a guiding principle in military R&D. Barnett envisions a battlefield where soldiers and drones operate as a single, interconnected organism—where a soldier’s eyes become targeting systems, and their voice becomes a weapon’s trigger.

From Sci-Fi to Battlefield Reality

The concept of soldiers using AR to enhance situational awareness isn’t entirely new. The U.S. military has experimented with head-mounted displays for decades, from the early 2000s Land Warrior program to more recent efforts like the Integrated Visual Augmentation System (IVAS), a Microsoft HoloLens-based project that ultimately faltered under scrutiny. But Anduril and Meta are taking a different approach—one that prioritizes speed, adaptability, and AI integration over bulky hardware and rigid interfaces.

Their first major project is the Army’s Soldier Borne Mission Command (SBMC), a $159 million prototyping contract awarded to Anduril in 2023. This initiative aims to develop AR glasses that clip onto existing military helmets, overlaying real-time data—like maps, drone feeds, and AI-identified threats—directly into a soldier’s field of view. Unlike previous systems that required complex menus or handheld controllers, Anduril’s design relies on natural interaction: voice commands and eye-tracking.

But Anduril isn’t putting all its eggs in one basket. Alongside SBMC, the company is developing EagleEye—a self-funded, proprietary helmet and headset combo designed from the ground up for military use. Unlike SBMC, which adapts commercial tech for defense, EagleEye is built specifically for combat, with ruggedized components, enhanced battery life, and deeper integration with Anduril’s proprietary Lattice software platform. Anduril believes the military will eventually prefer this custom solution, even if it wasn’t originally requested.

The Human-Machine Symbiosis

The true innovation lies not in the hardware, but in the software that powers it. At the core of both SBMC and EagleEye is Anduril’s Lattice platform—a neural network of sorts that fuses data from drones, satellites, ground sensors, and even other soldiers into a single, real-time operational picture. Imagine a soldier in a firefight who can see not only enemy positions, but also the flight paths of nearby surveillance drones, the status of friendly units, and even predictive AI models suggesting optimal routes or targets.

This level of integration represents a leap beyond traditional command-and-control systems. Instead of relying on radios, maps, and fragmented intel, soldiers receive contextualized, visual information that’s tailored to their immediate needs. A compass might appear when navigating, or a thermal signature of a hidden vehicle might glow red when a drone spots it. The system uses AI to filter noise and highlight only what’s relevant—reducing cognitive load and accelerating decision-making.

But the real game-changer is the interface. Anduril is testing large language models (LLMs) from Google (Gemini), Meta (Llama), and even Anthropic (Claude)—despite the latter’s public stance against military contracts—to enable natural language interaction. A soldier could say, “Evacuate the wounded at grid 483921,” and the system would automatically route a medical drone, alert nearby units, and update the mission log. Or, “Send a drone to scan the ridge to the north,” and a surveillance UAV would launch within seconds.

This voice-to-action capability is powered by AI that translates human speech into machine commands, bridging the gap between instinct and execution. It’s not just about convenience—it’s about survival. In high-stress environments, every second counts, and reducing the steps between observation and action can mean the difference between life and death.

Eye-Tracking: The New Trigger

Perhaps the most futuristic—and controversial—feature of Anduril’s AR glasses is eye-tracking. The system can detect where a soldier is looking and use that gaze as a targeting mechanism. Spot an enemy vehicle? Glance at it, and the system could automatically tag it for drone surveillance or even a strike. This “gaze-and-strike” capability echoes the kind of human-machine fusion seen in fighter jets, where pilots use helmet-mounted sights to lock onto targets simply by looking at them.

But unlike jet cockpits, where the environment is controlled and predictable, ground combat is chaotic. A soldier might glance at a civilian vehicle, a child, or a decoy—raising serious ethical and operational concerns. Anduril insists that eye-tracking will be used only for identification and targeting suggestions, not autonomous weapon release. Final decisions, the company says, will remain in human hands.

Still, the technology pushes the boundaries of what’s acceptable in modern warfare. Critics warn that even suggestive targeting could lower the threshold for violence or lead to errors in high-pressure situations. And with AI increasingly involved in threat recognition, the risk of misidentification—especially in complex urban environments—remains a significant challenge.

The Road to 2028 and Beyond

Despite the promise, both SBMC and EagleEye are still in the prototyping phase. The Army doesn’t expect to move its top choice into production until 2028—if it chooses one at all. The failure of Microsoft’s IVAS program serves as a cautionary tale: even well-funded, high-profile projects can collapse under the weight of technical limitations and operational realities.

Anduril’s advantage may lie in its agility. Unlike legacy defense contractors, Anduril operates with a startup mindset—rapid iteration, private funding, and a focus on software over hardware. Its self-funded EagleEye project, for example, allows it to experiment without waiting for government approval. This dual-track approach—working within the military’s framework while pursuing independent innovation—could give Anduril an edge in a field where speed is as critical as precision.

Moreover, Anduril’s partnership with Meta brings significant resources. Meta’s expertise in consumer AR—honed through products like Ray-Ban Stories and the Quest headset—provides a foundation for lightweight, user-friendly designs. While military applications demand far greater durability and security, the core principles of intuitive interaction and seamless integration remain the same.

Ethical Crossroads and the Future of War

As these technologies advance, they raise profound ethical questions. Who is responsible if an AI misidentifies a target and a drone strike kills civilians? Can a soldier truly maintain situational awareness when relying on AI-generated overlays? And what happens when adversaries develop similar systems, leading to an arms race in augmented warfare?

Anduril argues that its systems are designed to enhance, not replace, human judgment. “We’re not building killer robots,” Barnett has said. “We’re building tools that help soldiers make better decisions faster.” But the line between assistance and automation is thin—and blurring fast.

The military’s adoption of AR smart glasses could also reshape the very nature of combat. Future battles may be fought not just with bullets and bombs, but with data, algorithms, and networked awareness. The soldier of 2030 may be less a lone warrior and more a node in a vast, intelligent battlefield ecosystem.

Conclusion: The Soldier as a System

The collaboration between Anduril and Meta represents more than a technological upgrade—it’s a philosophical shift in how we think about warfare. The goal isn’t just to make soldiers stronger or faster, but to integrate them into a larger, smarter system where human intuition and machine intelligence operate in harmony.

Whether through the Army’s SBMC program or Anduril’s independent EagleEye project, the vision is clear: the future soldier will wear their command center on their face. And with it, they’ll carry the weight of decisions that could reshape the battlefield—and the world.

This article was curated from Inside Anduril and Meta’s quest to make smart glasses for warfare via MIT Technology Review