Unaligned Newsletter
Posts
AI in Autonomous Weapons

AI in Autonomous Weapons

Robert Scoble & Irena Cronin
May 27, 2025

The integration of AI into military technologies has fundamentally transformed the nature of modern conflict. While AI promises to optimize decision-making, enhance target precision, and reduce human casualties in combat, its use in autonomous weapons systems—machines capable of selecting and engaging targets without human intervention—has raised profound ethical, legal, strategic, and technological concerns.

Defining Autonomous Weapons

Autonomous Weapons Systems (AWS), sometimes referred to as “killer robots,” are weapons platforms that can identify, select, and engage targets without direct human control. These include unmanned aerial vehicles (UAVs), unmanned ground vehicles (UGVs), naval drones, and advanced missile systems.

While automation in military systems has existed for decades, such as guided missiles or semi-autonomous surveillance drones, the leap to fully autonomous decision-making—particularly regarding the use of lethal force—is unprecedented. At the heart of these systems lies AI, which enables real-time processing of vast amounts of data, environmental adaptation, and predictive modeling.

Key AI Capabilities in Autonomous Weapons

Computer Vision and Target Recognition
AI-powered systems can process visual and sensor data to recognize specific objects, vehicles, or individuals. Facial recognition, heat signatures, and movement patterns can be used to differentiate between civilians and combatants—or at least attempt to.
Pathfinding and Navigation
Algorithms allow autonomous drones or robots to navigate complex terrains, avoid obstacles, and optimize travel routes without GPS or human guidance.
Swarm Intelligence
Some systems employ decentralized AI coordination, allowing multiple units (e.g., drone swarms) to cooperate without central control. This enables rapid adaptation to battlefield dynamics.
Decision-Making Under Uncertainty
Reinforcement learning and probabilistic reasoning are used to make real-time choices in ambiguous or rapidly changing environments—decisions that could involve initiating lethal force.
Continuous Learning
Some systems are designed to adapt from previous engagements, adjusting tactics based on outcomes, making them increasingly efficient—and unpredictable.

Benefits of AI in Autonomous Weapons

Reduced Risk to Military Personnel

Removing humans from frontline combat can save lives by distancing soldiers from immediate harm. Autonomous systems can be deployed in high-risk scenarios such as minefields or urban warfare zones.

Faster Response Times

AI systems can analyze information and make decisions significantly faster than humans, allowing for split-second reactions in high-stakes situations.

Operational Efficiency

Autonomous weapons can operate continuously, without fatigue, across challenging conditions such as extreme temperatures or electromagnetic interference zones.

Precision and Minimization of Collateral Damage

Advanced targeting algorithms aim to improve accuracy, reducing unintended casualties—though this remains a contested claim.

Ethical and Legal Challenges

Delegating Lethal Decision-Making

The core ethical issue is whether machines should be granted the authority to make life-and-death decisions. Critics argue that such decisions require moral judgment that AI simply cannot replicate.

Accountability and Liability

If an autonomous system commits a war crime, who is responsible? The programmer? The manufacturer? The commander who deployed it? This diffusion of accountability is legally and morally troubling.

Compliance with International Humanitarian Law (IHL)

Autonomous weapons must adhere to IHL principles, including distinction (between combatants and civilians), proportionality, and military necessity. Given AI’s current limitations in contextual understanding, compliance remains questionable.

Risk of Malfunction or Exploitation

AI systems can malfunction or be hacked, leading to unintended engagements or catastrophic friendly fire incidents. Furthermore, adversaries might exploit known AI weaknesses through adversarial attacks or deception tactics.

Strategic and Geopolitical Implications

Arms Race Dynamics

The potential for strategic advantage is pushing major powers—including the U.S., China, and Russia—to accelerate development. This has sparked a modern AI-driven arms race with minimal international oversight.

Lowering the Threshold for Conflict

The reduced cost (both human and financial) of deploying autonomous weapons may make war more "palatable," encouraging states or non-state actors to engage in conflict more readily.

Asymmetric Warfare

Autonomous weapons could exacerbate power imbalances between technologically advanced nations and developing states. Conversely, relatively inexpensive drone swarms could empower insurgent groups or rogue states, complicating defense strategies.

Proliferation Risks

As AI technology becomes more accessible, autonomous weapons could proliferate to authoritarian regimes, terrorist organizations, and criminal enterprises—leading to destabilization and untraceable acts of violence.

The Case for Regulation and Control

Calls for a Ban or Moratorium

A growing number of scientists, ethicists, and civil society organizations—such as the Campaign to Stop Killer Robots—are advocating for preemptive bans on fully autonomous weapons. In 2015, a letter signed by over 1,000 AI researchers, including Stephen Hawking and Elon Musk, warned of the destabilizing effects of such systems.

UN Involvement

The United Nations Convention on Certain Conventional Weapons (CCW) has held ongoing discussions about the legality and regulation of AWS. However, progress has been slow and non-binding due to lack of consensus among major military powers.

“Human-in-the-Loop” Safeguards

Many experts support the idea that all lethal decisions should involve meaningful human control. This includes:

Human-in-the-loop: System suggests actions, but human initiates them.
Human-on-the-loop: System acts autonomously but can be overridden by a human.
Human-out-of-the-loop: No human intervention once deployed—this is the most controversial category.

The Future of AI-Powered Warfare

Human-Machine Teaming

Future battlefields may increasingly rely on mixed teams where autonomous systems work alongside soldiers, providing real-time intelligence, reconnaissance, and even decision-support in combat.

Algorithmic Warfare and Decision Support

Beyond physical combat, AI will continue to shape digital and strategic decision-making in command centers, predicting adversary moves, optimizing logistics, and deploying cyberattacks.

Autonomous Defense Systems

Countries are developing AI-powered defense platforms to detect and intercept incoming missiles, cyberattacks, or even drone incursions—expanding the AI arms race into the realm of national security infrastructure.

Dual-Use Dilemma

The same algorithms used in self-driving cars or robotic surgery can be repurposed for autonomous weapons. This dual-use nature makes monitoring and controlling proliferation even more complex.

AI in autonomous weapons is not just a technological issue—it is a humanitarian, philosophical, and geopolitical dilemma. While AI offers undeniable operational advantages, it also poses existential threats to the very fabric of warfare ethics and international stability. The world stands at a crossroads: we can either allow AI to accelerate an uncontrolled arms race, or we can choose to shape this technology with foresight, accountability, and humanity at the forefront.

As AI continues to redefine the boundaries of possibility, the question remains not just what these systems can do—but what they should do, and who gets to decide.

Just Three Things

According to Scoble and Cronin, the top three relevant and recent happenings

OpenAI Acquires Jony Ive’s io to Design Next-Gen AI Hardware

OpenAI is acquiring hardware startup io, founded by former Apple design chief Jony Ive and other ex-Apple engineers, in a deal valued at nearly $6.5 billion. While Ive’s design firm LoveFrom will remain independent, it will lead all of OpenAI’s software and hardware design efforts. About 55 engineers and product experts from io, including Evans Hankey, Scott Cannon, and Tang Tan, will join OpenAI. The first device—described as pocket-sized, screen-free, and contextually aware—is expected in 2026. Sam Altman and Ive have been collaborating for two years on this new category of AI-powered hardware, which they say is designed to enhance human potential without aiming to replace smartphones. The Verge

OpenAI and Jony Ive Reveal Plans for Screenless, AI-Powered Personal Companion

OpenAI and Jony Ive are developing a screenless, pocket-sized AI device designed to act as a personalized, context-aware companion that blends seamlessly into everyday life. Unlike smartphones or laptops, this device will focus on ambient interaction, aiming to reduce users’ dependency on screens. According to leaks from an internal OpenAI meeting, the product will function as a third core device alongside a MacBook and iPhone, and will be fully aware of the user’s surroundings. Drawing inspiration from the AI companion in the film Her, the project marks a bold attempt to redefine how people interact with technology—without relying on XR glasses or traditional interfaces. Mashable

Google Launches Veo 3: AI Video Generator with Integrated Audio and Lip Syncing

Google has launched Veo 3, a powerful AI video generator that sets itself apart from competitors like OpenAI’s Sora by incorporating synchronized audio—including dialogue and ambient sounds—into generated videos. Available to U.S. users via the $249.99/month Ultra subscription and Vertex AI, Veo 3 also supports accurate lip-syncing and real-world physics. Google simultaneously unveiled Imagen 4 for high-quality image generation and Flow, a cinematic filmmaking tool accessible through platforms like Gemini and Workspace. These announcements mark Google's latest push into multimedia AI, following mixed results from previous models and growing competition in the generative media space. CNBC