AI-Driven War Machines: Can They Be Trusted?

Tanks with military personnel in street parade formation.

The U.S. Army just showed the world its first look at the vehicles meant to replace the Bradley Fighting Vehicle — and what they revealed is both impressive and, if history is any guide, worth watching very carefully.

Story Snapshot

The Army publicly unveiled computer-aided design renderings of two competing XM30 infantry fighting vehicle candidates at the National Defense Industrial Association’s MDEX 2026 event in May 2026.
General Dynamics Land Systems and American Rheinmetall Vehicles are the two finalists, each awarded contracts worth a combined $1.6 billion in June 2023.
The XM30 is designed to replace the aging M2 Bradley with AI-assisted targeting, a hybrid electric powertrain, and significantly improved lethality, protection, and mobility.
Both contractors completed their Preliminary Design Reviews in August 2024 and have since advanced to detailed design — but no physical prototype has yet been publicly demonstrated.

The Bradley Has Been Waiting for a Successor for Decades

The M2 Bradley entered service in 1981. It has been upgraded repeatedly, fought in Desert Storm, Iraq, and most recently proven surprisingly effective against Russian armor in Ukraine. But the Army has known for years that the Bradley’s underlying architecture has limits. The XM30 Mechanized Infantry Combat Vehicle program exists because those limits are becoming harder to paper over with incremental improvements. The question is whether the replacement will actually arrive — and work — before the Bradley becomes a liability on a modern battlefield. ^[4]

At the National Defense Industrial Association’s MDEX 2026 conference, the Army released official computer-aided design renderings of both competing designs. American Rheinmetall’s entry, called the XM30 Lynx, draws on the company’s established Lynx platform already in service with other allied militaries. General Dynamics Land Systems brings its own design lineage and cleared its Critical Design Review, advancing the program into its third and most detailed phase. ^[1] Both designs incorporate modular open systems architecture, meaning future upgrades — including weapons, sensors, and software — can be swapped without rebuilding the vehicle from scratch. ^[3]

What AI Actually Means in This Context

The “AI” label attached to the XM30 deserves some unpacking. This is not a self-driving tank. The artificial intelligence components under development focus on decision-support tools — systems that help crews process battlefield data faster, prioritize threats, and reduce the cognitive load on soldiers already managing an overwhelming amount of information in combat. The vehicle is also being designed as optionally manned, meaning it can operate with or without a crew depending on the mission. That is a meaningful capability leap, not a marketing gimmick, though it remains unproven in operational conditions. ^[3]

The hybrid electric powertrain is equally significant. Electric drive systems offer quieter operation, faster torque response, and the ability to generate onboard power for energy-hungry electronic systems without running a separate generator. Modern warfare increasingly depends on sensors, communications, and electronic warfare equipment that all demand reliable power. A vehicle that can feed those systems while staying tactically quiet is genuinely more capable than one that cannot. ^[3]

Pretty Renderings Are Not Combat-Proven Vehicles

Here is where honest assessment matters. What the Army revealed at MDEX 2026 were computer-aided design images — engineering renderings, not hardware. ^[5] Defense acquisition programs have a long and expensive history of looking spectacular on screen and then encountering brutal reality during prototype testing. The Army’s own program timeline places prototype construction in fiscal year 2025, meaning physical vehicles are only now beginning to exist. ^[6] No survivability testing, no live-fire results, and no operational data have been made public. The program is progressing on schedule by acquisition standards, but acquisition standards and battlefield readiness are very different measures. ^[1]

The broader defense acquisition pattern here is worth naming plainly. The system rewards visible milestones: contract awards, design reviews, prototype announcements. Each milestone generates press coverage and congressional confidence. What the system does not reward — at least not until it becomes unavoidable — is confronting the gap between a polished rendering and a vehicle that can absorb a hit, keep its crew alive, and keep fighting. The XM30 program appears well-managed by current standards, but well-managed programs have still delivered late, over budget, and underperforming. Healthy skepticism is not cynicism — it is the lesson of every major ground vehicle program since the Cold War. ^[6]

Two Strong Competitors, One Serious Program

What separates the XM30 from some past failures is the competitive structure. Two credible contractors — one with deep American defense roots, one with a proven international infantry fighting vehicle platform — are driving each other toward better solutions. The Army also made a deliberate choice to use digital engineering tools throughout development, which allows designers to identify and fix problems in software before they become expensive metal mistakes. ^[1] That approach has worked well in aerospace programs and represents genuine modernization in how the Army buys ground vehicles. The $1.6 billion combined contract value signals serious institutional commitment, not a study program destined to be quietly cancelled. ^[3]

The Bradley served America’s soldiers for over forty years. Its successor deserves to be built right, not just built fast. The XM30 designs look promising. The real test — the only one that matters — comes when these vehicles stop being renderings and start taking fire.