The Future of Autonomous Weapons Systems

A preview of the next cold war

The Russia-Ukraine war has catalyzed rapid development in autonomous weapons systems. Militaries typically plan using doctrines learned from previous conflicts, adapting as new battlefield experiences dictate. Global military strategists are closely observing this conflict's evolution, while defense technology startups deploy their systems to Ukraine for real-world testing against Russian forces. By 2030, the "drone wars" are likely to evolve into a form of technological cold war between the US and China, with superpowers deploying vast numbers of autonomous systems in proxy conflicts.

Timeline of the Evolution of Autonomous Weapons Platforms:

Phase 1: Semi-Autonomous Systems and Short-Duration Loitering

These technologies are currently being tested in Ukraine:

• Long-range autonomous naval drones: Unmanned Surface Vessels (USVs) loaded with explosives. They utilize GPS and satellite communications for long-range navigation but can switch to AI-driven sensors for terminal guidance to overcome Electronic Warfare (EW) measures and identify specific targets.

• Aerial target acquisition and tracking: Unmanned Aerial Vehicles (UAVs) master locking onto targets, then independently track and engage the target, including both vehicles and individual combatants. This includes integration with fire control systems for both direct and indirect fire weapons.

• Multi-sensor target identification: UAVs and ground-based sensors independently identify and designate targets on the battlefield. A human operator validates the target, which is then engaged by networked weapons platforms such as artillery or loitering munitions.

• Short-duration loitering munitions: Aerial weapons systems that can loiter for several hours while awaiting opportunities. While similar to previous UAV designs, they can operate semi-autonomously and are significantly more cost-effective than traditional U.S. drone systems. Examples include the Switchblade and Phoenix Ghost systems.

Phase 2: Extended Autonomous Operations (2026)

• Persistent UAV systems: Rather than being deployed for specific missions, these UAVs are deployed in large numbers (hundreds to thousands) over a wide area. They can operate autonomously for weeks, awaiting targets of opportunity. Human operators validate targets, then authorize the UAVs to prosecute the target independently. These systems will incorporate AI-driven mission planning and dynamic re-tasking capabilities.

• Autonomous Intelligence, Surveillance, and Reconnaissance (ISR) systems: UAVs can remain airborne for months, either through autonomous recharging or advanced power systems like high-altitude solar-powered platforms. A network of these UAVs can provide continuous coverage over large operational areas.

• AI-enhanced target identification: Working with edge computing and distributed AI systems, aerial and ground-based platforms can identify and categorize potential adversaries for targeting. This includes behavioral analysis and pattern-of-life assessments to differentiate combatants from civilians.

• Autonomous launch and recovery systems: These systems can self-deploy from field stations with remote human authorization. Concealed remote launch sites can operate on the front lines while their human operators remain at safe distances. This includes both aerial and ground-based autonomous systems.

• Autonomous ground combat vehicles: Tracked and wheeled weapons platforms capable of navigating complex terrain autonomously, acquiring targets, and engaging after human approval. These will include modular designs that can be rapidly reconfigured for different mission profiles.

Phase 3: Integrated Autonomous Battle Networks (2028)

Multi-domain autonomous systems: The integration of various autonomous platforms, including:

• Persistent aerial ISR and strike platforms

• Autonomous launch and recovery systems for various munitions

• Unmanned ground combat vehicles

• Naval autonomous systems (surface and subsurface)

These systems provide commanders with (1) a comprehensive, real-time operational picture and (2) distributed, resilient deployment capabilities. While human supervisors still oversee each system type, increasing automation allows a single operator to control multiple platforms across domains. At this stage, human input is primarily required for designating operational areas, confirming Rules of Engagement (ROE) compliance, and authorizing kinetic actions.

• Mass production of autonomous platforms: Once their efficacy is proven and designs are standardized, industrial capacities are leveraged to mass-produce autonomous systems. Major military powers develop the capability to field millions of autonomous systems, deploying them in proxy conflicts for further refinement and validation. This will likely lead to the development of counter-autonomous warfare doctrines and technologies.

• AI-driven systems replace human-in-the-loop control: First-Person View (FPV) systems, which use VR headsets for drone control, become obsolete as the operational tempo outpaces human reaction times. Air, land, and sea-based autonomous systems now only require human oversight for ROE compliance and strategic decision-making, with all tactical decisions and engagements managed by AI.

Phase 4: Full-Spectrum Autonomous Warfare (2030)

By this point, the battlefield has become inhospitable to human combatants due to the prevalence of autonomous systems:

• Precision autonomous weapons have nullified traditional tactics like suppressive fire and static defense.

• Advanced sensor fusion and predictive AI render physical camouflage and concealment largely ineffective against swarms of networked autonomous platforms.

• Persistent autonomous ISR and weapons systems can identify and engage targets continuously, making unprotected human movement in contested areas extremely hazardous.

Given these developments, military strategy evolves to prioritize:

• Mass deployment: Millions of autonomous systems are fielded, with emphasis on overwhelming adversary defenses through sheer numbers and distributed lethality.

• Advanced AI integration: Systems increasingly determine their own operational areas, identify adversaries, and engage targets with minimal human input, guided by strategic-level commands and ROE.

• Edge computing supremacy: Comprehensive electronic warfare and anti-satellite capabilities deny access to remote compute resources. Therefore, battlefield advantage goes to the side with superior local, hardened computing capabilities.

• Mesh network resilience: Compute resources dynamically allocate across local battle networks, with AI optimizing processing distribution to support engaged systems and maintain network integrity under attack.

Autonomous area denial systems: Highly autonomous weaponized platforms designed to control large areas by engaging any unauthorized presence. These may target critical infrastructure operators, command and control nodes, and civilian populations as part of broader strategic operations.

Countering autonomous systems: With engagement speeds far exceeding human reaction times, only AI-driven countermeasures provide effective defense. Directed energy weapons, electromagnetic pulse (EMP) systems, and dedicated counter-drone platforms are deployed en masse to protect vital assets, infrastructure, and population centers. Cyber and electronic warfare capabilities become critical in disrupting enemy autonomous networks.