Algorithms to Ahimsa: Building India’s AI Warfare Ethics Doctrine
| AUTHOR | Prajakta Pawar |
| DATE | September 30, 2026 |
| CATEGORIES | AI Defence |
Autonomous weapons have moved from prototype to daily use. The shift toward machine-initiated action is significant for accountability. The technology is also moving towards greater autonomy — reports from Ukraine describe AI-driven drone swarms that can coordinate among themselves and automatically redistribute when individual drones are destroyed. At the same time, the global market for autonomous weapons systems, valued around $14 billion in 2024, is expected to reach $33.14 billion by 2034. This suggests that decisions on warfare are increasingly being made at machine speed, and these technologies are arriving faster than the rules meant to govern them.
The Accountability Gap
That speed creates a specific moral problem. When an algorithm identifies a target and a human confirms the strike within seconds, the legal responsibility formally rests with the authorising commander. However, automation bias is a more relevant issue that is minimised by this legal clarity. Studies on human-machine decision-making show that operators presented with a confident algorithmic recommendation tend to approve it — often without the same level of questioning they would apply to another human recommendation — even when they technically have the authority to reject it. The responsibility diffuses and spreads across the programmer, the manufacturer, the commander, and the operator, where each contributes to the final act but none bears the full responsibility. When the system misreads its target — a civilian, critical infrastructure, or a friendly unit — it becomes difficult to determine who should be held accountable. As decision-making is shared between humans and machines, the question of who should be held responsible becomes harder to answer, not easier.
Operation Sindoor and the Governance Lag
India provides a useful case study. During Operation Sindoor in May 2025, the Akashteer air-defence system brought together data from multiple radars and sensors to detect and intercept aerial threats in real time. Lt Gen Rajiv Kumar, Director General of the Indian Army’s Electronics and Mechanical Engineers Corps, later explained that AI-assisted fusion helped in identifying and prioritising targets. Though the operation showed the military value of AI-enabled systems, it also pointed to a challenge: as AI becomes part of the targeting process, it raises the question of accountability.
The US Department of Defense’s Directive 3000.09 requires “appropriate levels of human judgment” over lethal autonomous systems, and its 2020 AI Ethical Principles commit to systems being responsible, equitable, traceable, and governable. Meanwhile, India’s initiatives such as the National AI Strategy, iDEX, and the Defence AI Council have focused on building capabilities, but ethical governance has received far less attention. A parliamentary committee report tabled in the Lok Sabha in March 2026 recorded the Defence Ministry’s own concern that AI/ML techniques are “not amenable for verified decision making” and can produce unintended outcomes, even as DRDO continues active work on lethal autonomous weapon systems. India has already recognised the gap — the next step needed is an AI ethics doctrine that places accountability alongside capability.
A Decade of International Deadlock
The international efforts to regulate autonomous weapons have been running for over a decade without resolution. States have discussed lethal autonomous weapons at the UN’s Convention on Certain Conventional Weapons since 2014, formalised as a Group of Governmental Experts from 2016, with the current mandate closing at the Seventh Review Conference in November 2026. The US and Russia have resisted a binding framework; China has supported broad principles even as it expands its own autonomous weapons programme. UN General Assembly Resolution 79/62, passed with 166 votes in December 2024, raised concerns about autonomous weapons — especially the risk of arms races, escalation, and their spread to non-state actors — and called for consultations among all member states to explore possible regulations.
India chose to abstain, arguing that discussions on LAWS should occur within the GGE framework under the CCW, emphasising consensus rather than parallel or rushed treaty processes. Even so, a legally binding global treaty remains distant. This deadlock gives India an opportunity to develop an independent national doctrine.
Classical Traditions as Ethical Foundation
Under customary international humanitarian law, Rule 14 of the ICRC’s Customary IHL Study codifies proportionality — prohibiting attacks expected to cause civilian harm that outweighs the anticipated military advantage. It is a legal threshold that a commander must assess before authorising a strike.
India’s classical traditions offer useful ideas for thinking about AI ethics in warfare — not as rules to be applied literally, but as ethical frameworks. Ahimsa, drawn from Jain, Buddhist, and Gandhian thought, sits alongside international humanitarian law as a distinct but complementary ethical idea: a principle of non-violence and restraint. The Mahabharata’s Dharma-yuddha stresses the protection of non-combatants, resonating with modern rules of precaution and target verification. Similarly, the Arthashastra’s distinction between Niti (procedural correctness) and Nyaya (substantive justice) highlights an accountability gap that exists across autonomous systems generally — that a system can follow its rules perfectly and still produce an unjust outcome. This offers a way to think about accountability when AI systems cause unintended harm. And finally, Danda, the calibrated use of coercive force, gives a useful lens for managing escalation in an era of machine-speed warfare.
The Case for a National Doctrine
Together, these traditions offer India a distinct ethical foundation to build with. The need extends beyond India’s borders. The CCW has debated for over a decade without producing binding rules, largely because the states writing them are also the states developing the technologies they seek to regulate. A well-reasoned and transparent national doctrine could itself be a contribution to that debate — demonstrating that principled regulation of autonomous weapons is achievable. By developing its own framework, India would be able to contribute ideas and principles to the global debate, if it pairs philosophical tradition with an explicit account of what national security interests are served by principled regulation, and what risks accumulate when regulation is absent.