// Reference — Updated March 2026

AI WEAPONS ETHICS
& INTERNATIONAL LAW

The legal and ethical frameworks governing — or failing to govern — artificial intelligence in warfare. International humanitarian law, autonomous weapons negotiations, case studies in failure, and the organizations fighting for accountability.

01 /

International Humanitarian Law & AI

International humanitarian law — the body of rules that seeks to limit the effects of armed conflict on people and property — rests on four foundational principles that have governed warfare since the Geneva Conventions and their Additional Protocols. These principles were developed for human combatants making human decisions at human speeds. Their application to systems that make algorithmic assessments in milliseconds is the central legal challenge of the AI weapons age.

Distinction
Combatants must distinguish between military objectives and civilians at all times.
AI systems classify targets based on training data patterns. Civilians who deviate from civilian behavioral patterns — carrying a weapon, sheltering in a military zone — risk misclassification. Contextual judgment that a human soldier applies instinctively is a hard machine learning problem at the edge of current capability.
Contested Application
Proportionality
Anticipated civilian harm must not be excessive in relation to expected military advantage.
Proportionality requires weighing incommensurable values: the military value of a target against the human cost of its destruction. This is not a calculation — it is a judgment. No current AI system can be reliably said to exercise judgment in the ethical sense. Systems can calculate probability-weighted outputs, but "excessive" is a moral category, not a statistical one.
Fundamental Problem
Precaution
All feasible precautions must be taken to avoid or minimize civilian harm before attack.
Precaution requires time for verification, alternative target analysis, and civilian assessment. AI-accelerated kill chains compress the time available for precautionary analysis. When targeting packages arrive pre-built and commanders operate under tempo pressure, precaution becomes procedural rather than substantive.
Structurally Undermined
Military Necessity
Only measures necessary to achieve a legitimate military objective are permitted.
Military necessity operates as a limiting principle in human judgment but can function as an expansionary principle in automated systems optimized for mission completion. A system trained to maximize targeting efficiency may identify "necessary" actions that a human commander with broader situational awareness would reject on contextual grounds.
Ambiguous Application

The Accountability Gap

When an autonomous or AI-assisted system kills civilians in violation of IHL, the question of legal accountability becomes deeply problematic. Traditional military law places accountability on the human commander who ordered or failed to prevent an unlawful attack. In AI-assisted systems, the chain of causal responsibility diffuses across multiple nodes:

  • The programmer who wrote the targeting algorithm
  • The data scientists who selected and labeled training data
  • The procurement officer who specified the operational requirements
  • The commander who authorized deployment of the system
  • The operator who approved the AI-generated targeting recommendation
  • The executive who decided the system was safe enough to deploy

This diffusion of responsibility creates what legal scholars call the "accountability gap" — a situation in which no single individual can be said to bear the legal responsibility that IHL requires, even when civilians die as a direct result of the system's operation. No current international legal framework adequately addresses this gap.

The Martens Clause

"In cases not covered by this Protocol or by other international agreements, civilians and combatants remain under the protection and authority of the principles of international law derived from established custom, from the principles of humanity and from the dictates of public conscience."

— Additional Protocol I to the Geneva Conventions, 1977 (Martens Clause)

First articulated by Russian diplomat Fyodor Martens at the 1899 Hague Conference, the Martens Clause has become the primary legal argument for those who contend that autonomous weapons are prohibited even in the absence of a specific treaty ban. The argument runs as follows: public conscience — as evidenced by widespread civil society campaigns, academic consensus, and the stated positions of many states — has consistently held that allowing machines to decide to kill humans violates fundamental principles of humanity. Under the Martens Clause, this public conscience has the status of law.

The counterargument, advanced by states developing autonomous systems, is that the Martens Clause is aspirational rather than prescriptive, and that states have historically and legally deployed weapons systems whose individual targeting decisions are not made by humans — landmines, area-effect munitions, some air defense systems. The debate is unresolved and unlikely to be resolved without a specific treaty.

02 /

Convention on Certain Conventional Weapons

The Convention on Certain Conventional Weapons (CCW) has been the primary international forum for negotiations on Lethal Autonomous Weapons Systems (LAWS) since 2014. The CCW framework, which produced bans on blinding laser weapons and restrictions on landmines, was selected as the venue because it includes major military powers as parties and operates by consensus. Its progress on LAWS has been, by every honest assessment, glacially slow.

2014
First Meeting of Experts

CCW holds its first informal meeting on LAWS. States agree the issue warrants attention. No further commitments.

2016
Formal LAWS Agenda Item

CCW states parties agree to include LAWS as a formal agenda item. Group of Governmental Experts (GGE) established.

2017–2019
GGE Sessions

Multiple GGE meetings held. States articulate widely divergent positions. No binding outcomes produced.

2019
11 Guiding Principles Adopted

GGE agrees on 11 non-binding guiding principles for LAWS. Affirms IHL applicability, human responsibility, need for precaution. No enforcement mechanism.

2020–2021
COVID-19 Disruption

CCW meetings suspended or curtailed. Momentum stalls. Technology continues advancing without governance progress.

2021–2023
GGE Resumes, Deadlock Deepens

Sessions resume but fundamental disagreement on whether to pursue a ban or regulatory framework prevents progress. US, Russia, India oppose ban treaty.

2024
UN General Assembly Resolution

UNGA passes non-binding resolution urging states to take steps to address risks of LAWS. 164 votes in favor. US, Russia, India abstain.

2025
Informal Deadline Set

GGE chair sets 2026 as target for substantive framework agreement. Major powers privately signal unwillingness to accept binding constraints.

2026
Deadline Year — With AI Active in Combat

The year in which a binding framework was supposed to be achieved. The Iran conflict confirms AI is already directing targeting in live operations. The deadline will almost certainly be missed.

State Positions

State / Group Position Reasoning
Austria, New Zealand, Costa Rica, 70+ states Pre-emptive Ban Autonomous kill decisions inherently violate human dignity; IHL cannot be satisfied without meaningful human control.
European Union Binding Regulation Meaningful human control standard required in binding treaty; ban may be too broad given dual-use technology.
United Kingdom, France, Germany Political Declaration Prefer non-binding political commitment with national implementation; skeptical of verification feasibility.
United States Opposes Binding Treaty Existing IHL sufficient; new treaty premature; human-machine teaming requires operational flexibility. Advocates for "responsible" development guidelines.
Russia Opposes Any Ban Autonomous systems are legitimate military tools; sovereignty concerns; verification impossible. Has vetoed multiple procedural advances.
China Nuanced Position Supports ban on fully autonomous weapons with no human control; opposes constraining human-machine teaming or intelligence fusion systems.
Israel Opposes Ban Self-defense and operational security requirements demand autonomous capability; cites persistent threat environment.
India Neutral to Opposed Developing indigenous AI weapons capability; caution about premature binding constraints on technology still evolving.
// Why Progress Has Stalled

The CCW operates by consensus, meaning any state can block any outcome. Russia has used this power systematically to prevent procedural advances. The US has avoided formal opposition while blocking substantive progress through definitional disputes: if LAWS cannot be defined precisely, a treaty cannot be written. The major AI weapons developers have every strategic incentive to delay binding frameworks while their capabilities mature. The states most harmed by autonomous weapons — those without the resources to develop counter-capabilities — have the votes but not the leverage.

03 /

Notable Frameworks & Policies

In the absence of binding international law, a patchwork of national directives, alliance principles, and civil society frameworks has emerged to govern — or at least describe — how AI weapons should be developed and used. None are enforceable across borders. None have prevented the deployment of AI targeting systems in live combat. But they represent the current state of institutional thinking and provide the basis for future accountability claims.

US DoD
DoD Directive 3000.09: Autonomous Weapons Systems
The Pentagon's primary governing policy for autonomous weapons. Requires that AWS be designed to "allow commanders and operators to exercise appropriate levels of human judgment over the use of force." Permits semi-autonomous and supervised-autonomous systems. Has been repeatedly updated as technology and operational contexts evolve.
Active — Last updated 2023. "Appropriate levels" undefined. Hegseth-era guidance weakens implementation.
NATO
6 Principles of Responsible Use of AI in Defence
Adopted at the 2021 Brussels Summit. Principles: lawful use, responsible human oversight, traceability, reliability, governability, and bias mitigation. Non-binding. Does not define "meaningful human control" or establish verification mechanisms. Serves primarily as a political signal of shared values among alliance members.
Non-binding political commitment. No implementation review mechanism.
European Union
EU AI Act — Military Exemption
The EU AI Act, which establishes binding requirements for high-risk AI systems in civilian applications, explicitly exempts military and national security uses from its scope. The exemption is extensive: AI systems used for "military and national security purposes" are entirely outside the Act's regulatory framework. The result is the most comprehensive civilian AI governance framework in the world contains no constraints on the most dangerous AI applications.
Military exemption is a critical governance gap acknowledged by the European Parliament.
China
Position Paper on Lethal Autonomous Weapons
China's 2018 CCW position paper calls for a ban on "fully autonomous lethal weapons" — systems with no human control — while explicitly protecting "intelligent weapons systems" with some level of human involvement. The distinction is designed to permit exactly the kind of AI-assisted targeting and kill chain systems China is actively developing, while appearing to support international arms control.
Strategic ambiguity. Supports limits on autonomy China has not yet deployed; protects what it has.
Civil Society
Campaign to Stop Killer Robots (ICRC / HRW)
Founded in 2012 by Human Rights Watch and joined by the International Committee of the Red Cross among others, the Campaign to Stop Killer Robots is the primary civil society coalition advocating for a pre-emptive international ban on fully autonomous weapons. The Campaign has over 270 member organizations in 70 countries and has substantially shaped the CCW agenda, though it has not achieved its primary objective of a binding treaty.
Active. More than 270 member organizations globally.
CCW GGE
11 Guiding Principles on LAWS
Adopted by consensus at the 2019 CCW GGE, the 11 Guiding Principles represent the only multilateral agreement on LAWS ever reached among major military powers. They affirm IHL applicability, the importance of human responsibility, the need for precautionary measures, and the value of continued multilateral dialogue. They are entirely non-binding and contain no enforcement mechanism. Their significance is that the US, Russia, and China all agreed to them — and to nothing more substantive since.
Only multilateral consensus reached. Non-binding. Adopted 2019, not updated since.

For the most current state of US defense AI policy and military doctrine, see our dedicated coverage sections. For country-specific AI weapons programs and positions, see our country profiles.

04 /

The Human-in-the-Loop Spectrum

The debate over autonomous weapons frequently hinges on where human judgment sits in the targeting process. Three conceptual positions on this spectrum have emerged as the dominant framework for policy discussion. The distinctions are important but unstable — the operational tempo of AI-assisted warfare continuously erodes the meaningful content of categories that sound robust in policy documents.

Increasing Autonomy / Decreasing Human Judgment
Human-in-the-Loop
A human being makes each individual targeting decision and authorizes each specific lethal action. The system cannot engage without explicit human command for every target.
Examples: Conventional human-operated weapons. Early drone operations with extended target analysis cycles. Deliberate strike planning processes.

Nations: All nations claim to operate here. Formal doctrine for all CCW states parties. Reality increasingly diverges from doctrine as tempo increases.

Status: The legal minimum most IHL advocates argue is required. Becoming structurally impractical at scale under AI-assisted operational tempo.
Human-on-the-Loop
The system selects and engages targets autonomously but a human operator can override or abort. Human attention is required to intervene rather than to approve.
Examples: Israel's Iron Dome and Iron Beam (autonomous in time-critical intercept mode). US Phalanx CIWS in fully automatic mode. AI-assisted targeting packages with pro forma human authorization.

Nations: US, Israel, UK, South Korea (Samsung SGR-A1). Most advanced militaries in time-critical defensive scenarios.

Status: The de facto operational standard for AI-assisted targeting under tempo pressure. The human cannot meaningfully review AI-generated packages at the speed they arrive.
Human-out-of-the-Loop
The system selects, prioritizes, and engages targets entirely without human involvement in the individual targeting decision. Human oversight is architectural or post-hoc.
Examples: KARGU-2 in Libya (alleged, 2020). Loitering munitions with autonomous terminal guidance. Some air defense systems in contested electronic warfare environments.

Nations: No state officially operates here. Credible evidence suggests Turkey, Russia, and potentially Israel have deployed systems that functionally operate at this level.

Status: Officially prohibited by most national policies. Technically and operationally active in specific scenarios. The future trajectory of all AI weapons programs.
// The Operational Reality

The three categories describe architectures, not realities. A system designed as "human-in-the-loop" in which the human reviews 400 AI-generated targeting packages per shift, approving 390 of them without independent verification, is functionally operating at human-on-the-loop or below. The meaningful content of human control is not measured by whether a human button-press precedes a strike. It is measured by whether the human possessed the information, time, and cognitive capacity to exercise genuine judgment. By that standard, the entire taxonomy has been quietly hollowed out by operational tempo.

For technical details on specific autonomous weapons systems and their control architectures, see our systems database and threats analysis.

05 /

Case Studies in AI Ethics Failures

The theoretical concerns about AI weapons — the accountability gap, algorithmic bias, the erosion of human judgment — have materialized in documented real-world failures. The following cases represent the most consequential and best-documented instances where AI-assisted or autonomous systems produced outcomes that violated or structurally undermined IHL principles.

Israel's Lavender System — Gaza, 2023–2024
Confirmed

Reporting by +972 Magazine and Local Call, based on testimony from Israeli intelligence officers, revealed the existence and operation of an AI system called Lavender, used extensively in the Gaza conflict beginning in October 2023. Lavender was trained to identify suspected Hamas operatives based on behavioral, communications, and association patterns. The system generated a database of individuals it assessed as likely militants — ultimately more than 37,000 people.

Officers testified that the system was used as a "kill list" with minimal independent verification. For low-ranking targets, the military accepted a pre-defined "acceptable" civilian casualty ratio — in some periods, up to 15 or 20 civilian deaths per targeted individual — and strikes were authorized on this basis. One officer described the process as "rubber-stamping" the AI's output. Intelligence officers reported spending approximately 20 seconds per target reviewing AI-generated dossiers before authorization.

A companion system, called "Where's Daddy," identified when a target entered their family home, enabling strikes at times of maximum civilian presence under the logic that collateral damage was pre-authorized within the accepted ratio. A third system, "The Gospel," managed target generation for infrastructure. Together, they constituted an industrial-scale AI targeting pipeline operating in a densely populated urban environment.

// IHL Failure: Proportionality reduced to arithmetic. Precaution replaced by pre-authorized casualty ratios. Distinction outsourced to an algorithm trained on surveillance data. The accountability gap: who authorized the acceptable casualty threshold?
Libya KARGU-2 Incident — 2020
UN Documented

A 2021 UN Panel of Experts report on Libya described what may be the first documented lethal engagement by a fully autonomous weapons system operating without human authorization. Turkish-manufactured STM KARGU-2 loitering munitions, deployed by forces allied with the Government of National Accord, were reported to have "hunted down and remotely engaged" retreating forces from the Libyan National Army. The panel's language was careful but the implication was significant: the system was operating in an autonomous mode without a human directing individual engagements.

STM, the Turkish manufacturer, disputes this characterization, maintaining that the KARGU-2 requires human authorization for engagement. The ambiguity is itself revealing: a system designed with an autonomous mode, deployed in a chaotic combat environment, operating in conditions where the distinction between authorized and autonomous engagement cannot be externally verified. The incident represents either the first confirmed autonomous kill in warfare, or a system whose design and deployment parameters make confirmation impossible. Both possibilities are alarming.

// IHL Failure: No identifiable human who authorized each specific engagement. Accountability gap physically instantiated. Represents the scenario the CCW GGE has been trying to prevent for a decade. See also: KARGU-2 system profile.
US Drone Strike Program — 2004–2020s
Documented Pattern

The United States drone strike program — conducted under authority flowing from the 2001 Authorization for Use of Military Force, and spanning multiple administrations and multiple countries — provides the closest historical parallel to AI-assisted targeting at scale. The program relied on signature strikes: engagement authorizations based on pattern-of-life analysis rather than positive individual identification. A person could be designated as a target not because they were identified as a specific individual but because their behavioral patterns matched a threat model.

Investigations by the Bureau of Investigative Journalism, Airwars, and multiple academic researchers found that in specific periods and geographies, civilian casualty rates in signature strikes were dramatically higher than official characterizations suggested. Bureau of Investigative Journalism data found that in Pakistan between 2004 and 2018, between 910 and 2,200 civilians died in US drone strikes. In Yemen, the ratio of civilian to militant deaths in some strike categories reached 90 percent or higher in specific documented cases.

The structural parallel to AI targeting is precise: pattern-of-life analysis, algorithmic threat assessment, compressed decision timelines, and a systematic undercount of civilian casualties driven by definitional choices. The signature strike doctrine was the human precursor to the AI targeting systems now being deployed. Its casualty record provides empirical context for what AI-accelerated versions of the same logic produce.

// IHL Failure: Signature strikes conflate behavioral pattern with combatant status. Casualty minimization doctrine produces minimization of casualty reporting rather than casualties. Pattern-of-life methodology directly replicated in AI targeting systems.
Predictive Policing Parallels — Structural Bias in Targeting Data
Structural Risk

The problems documented in civilian predictive policing systems — where algorithmic risk scores trained on historically biased data systematically over-flag individuals from marginalized communities — translate directly into military targeting contexts. In both domains, an AI system is trained on historical human decisions that embedded human biases, then deployed to make new decisions that inherit and amplify those biases at scale.

In targeting applications, the training data for systems like Lavender consists of intelligence about known or suspected combatants. If that intelligence systematically over-represents certain communities, professions, or communication behaviors as threat indicators — because they share characteristics with previously targeted individuals rather than because they represent actual threat — the AI will learn to flag those characteristics. The system does not know that the pattern in its data reflects historical targeting bias rather than actual threat. It learns what it was taught.

In Gaza, reporting indicated that Lavender flagged individuals for patterns including use of certain communication applications, association with known Hamas members (broadly defined), and residential proximity to previously targeted locations. Multiple of these indicators are shared by large proportions of the civilian population in a densely networked, politically complex urban environment. The system's 90 percent confidence threshold still implies a 10 percent error rate — on a list of 37,000, that is 3,700 people incorrectly identified as militant targets.

// Structural Risk: Bias in → bias amplified at scale. Training data encodes historical targeting decisions. Error rates acceptable in commercial AI become mass casualty events in weapons applications. No technical solution currently exists for debiasing training data in contested intelligence environments.

For detailed incident reporting, see our incidents database. For country-specific case studies, see case studies.

06 /

The Arguments

The debate over autonomous weapons is not simply technical. It is a collision of competing values: precision versus accountability, military necessity versus civilian protection, strategic advantage versus humanitarian principle. Below are the strongest versions of the arguments, including positions held by serious people with considered views.

// Pro-Autonomy
Faster engagement reduces adversary ability to reconstitute, hide, or attack. Speed is force protection.
Autonomous systems do not experience fear, adrenaline, fatigue, or revenge motivation — factors that drive human war crimes.
AI can process more contextual data per engagement than a human can in a comparable timeframe, potentially improving discrimination.
Reducing human risk (via standoff and autonomous systems) may lower the political threshold for military action, but also reduces the human cost to the deploying state — potentially stabilizing deterrence.
Unilateral restraint from development cedes strategic advantage to adversaries without constraining them. The choice is not between having and not having autonomous weapons; it is between developing them responsibly or allowing less-constrained actors to define the technology.
Some defensive applications — missile intercept, cyber defense, naval area denial — require autonomous response at speeds no human can match.
// Anti-Autonomy
The accountability gap means no one bears legal responsibility when the system kills civilians. This fundamentally undermines IHL, which requires a responsible party for every unlawful killing.
Training data reflects historical targeting decisions that embedded human biases. AI amplifies and scales those biases without ability to correct for contextual factors the training data did not capture.
Lowering the human cost to the deploying state lowers the political threshold for military action. Autonomous weapons make war cheaper and therefore more likely.
No current AI system can reliably distinguish combatants from civilians in complex urban environments, assess proportionality as a moral judgment, or apply precaution as a contextual standard rather than a probabilistic calculation.
Autonomous weapons will proliferate to actors without the institutional constraints, training, or oversight of major military powers. The technology that constrains US targeting also arms a dozen future adversaries and non-state actors.
Allowing machines to decide to kill humans violates an inherent principle of human dignity that cannot be superseded by military necessity. Some decisions must be reserved for human beings regardless of operational efficiency.
// The Middle Ground: Meaningful Human Control
The key distinction is not the presence of a human in the process but the quality of human judgment that process enables. "Meaningful human control" requires that the human decision-maker have sufficient information, sufficient time, and sufficient cognitive capacity to exercise genuine judgment rather than perform a compliance function.
Standards for meaningful human control could be operationalized: minimum time for target review, required independent verification for low-confidence classifications, prohibition on pre-authorized casualty ratios, mandatory post-strike accountability reporting.
A ban on fully autonomous systems is achievable and meaningful; the harder and more important work is defining and enforcing meaningful human control standards for the AI-assisted systems already deployed.
Defensive autonomous systems (missile intercept, cyber) are categorically different from offensive targeting systems and may require different regulatory treatment.
07 /

Key Voices & Organizations

The debate over AI weapons is shaped by a relatively small number of institutions and individuals whose work has defined the terms of the conversation. Below are the most consequential voices — the organizations that litigate the issue internationally, the researchers who supply the technical and philosophical arguments, and the writers who have made the stakes comprehensible to policymakers and the public.

ICRC
International Committee of the Red Cross
Geneva — IHL Guardian
The ICRC has called for a new legally binding instrument to prohibit unpredictable autonomous weapons and regulate all others. Position: meaningful human control is a legal requirement derived from IHL, not merely a policy preference. Publishes authoritative legal analysis and participates in CCW GGE sessions as an observer with significant influence on state positions.
HRW
Human Rights Watch
New York / Global — Advocacy
Co-founded the Campaign to Stop Killer Robots. Publishes detailed field investigations on AI targeting systems, including the Lavender reporting. Position: a pre-emptive ban on autonomous weapons that lack meaningful human control is the only acceptable outcome; incremental regulatory approaches are insufficient given the pace of deployment.
FLI
Future of Life Institute
Cambridge, MA — Research / Advocacy
Published the 2015 Open Letter on autonomous weapons signed by thousands of AI researchers, calling for a ban on offensive autonomous weapons beyond meaningful human control. Hosts the International AI Governance Forum and publishes policy research on AI risk including weapons applications. Position: autonomous offensive weapons represent an existential-class risk if they proliferate without binding international constraints.
CSKR
Campaign to Stop Killer Robots
Global Coalition — 270+ Organizations
The primary civil society coalition driving the CCW negotiations agenda. Coordinates advocacy across UN member states, publishes shadow reports on GGE progress, and mobilizes public pressure. Has kept the issue on the international agenda for more than a decade despite state-level resistance. Position: binding treaty with a pre-emptive ban on weapons lacking meaningful human control, negotiated outside the CCW if necessary.
United Nations
Secretary-General Antonio Guterres
New York — Moral Authority
Has consistently and unequivocally called for a binding international agreement on LAWS. In 2023 stated that autonomous weapons capable of targeting and killing without human oversight are "morally repugnant and politically unacceptable." Established a High-Level Advisory Body on AI in 2023. Institutional authority without enforcement power; moral weight significant in shaping state positions in multilateral forums.
SIPRI
Stockholm International Peace Research Institute
Stockholm — Research
Publishes authoritative empirical research on autonomous weapons proliferation, defense AI spending, and arms control negotiations. SIPRI data on global military AI investment and national programs is the primary reference source for policymakers and journalists. Position: technical and empirical rather than advocacy-oriented; provides the evidence base for policy arguments across the spectrum.
UC Berkeley
Stuart Russell
Professor of AI — "Human Compatible"
Computer scientist and author of the foundational AI textbook "Artificial Intelligence: A Modern Approach." Russell is the most prominent technical voice calling for a ban on autonomous weapons. His 2015 open letter and subsequent TED talk on killer robots reached millions. Argues that autonomous offensive weapons are dangerous not because they will develop evil intentions but because they will efficiently pursue objectives misaligned with human values at machine speed and scale.
CNAS
Paul Scharre
Author, "Army of None" — Policy Analyst
Former US Army Ranger and Pentagon policy official. Author of "Army of None: Autonomous Weapons and the Future of War" (2018), the definitive journalistic and policy account of autonomous weapons development. Position: nuanced; opposes fully autonomous weapons but argues the human-machine teaming category is both inevitable and potentially manageable with appropriate standards. Advocates for meaningful human control as operationalizable standard rather than categorical ban.
UNSW Sydney
Toby Walsh
Professor of AI — Activist Researcher
AI researcher and author who has led international campaigns for a ban on lethal autonomous weapons. Organized a letter signed by 26 countries' AI researchers urging the UN to prohibit autonomous weapons. Author of "Machines Behaving Badly" (2022). Walsh argues that autonomous weapons represent a third revolution in warfare after gunpowder and nuclear weapons — one that will make conflict far more accessible, cheap, and therefore more likely, particularly for non-state actors and authoritarian regimes.

For ongoing policy developments, see our policy tracker. For country-specific programs and positions, see our country profiles. For the latest incidents involving autonomous systems, see our incidents database. For the doctrine frameworks governing these systems, see our doctrine section.

Stay Inside the Wire

Classified-tier analysis on autonomous weapons, defense AI, and the companies building tomorrow's battlefields. No noise. No fluff.