AI in the military: Path to ethical and strategic leadership

This article is authored by Zain Pandit, partner and Aashna Nahar, associate, JSA Advocates and Solicitors.

With defence emerging as one of AI’s most sensitive applications, a blend of innovation and responsibility is vital. India is laying the groundwork for responsible use of AI in defence, which is being led by key institutions like the Defence Research and Development Organisation (DRDO), NITI Aayog, and ministry of electronics and IT (MeitY).

AI(Unsplash)

To ensure ethical oversight, the DRDO launched the ETAI framework for evaluating trustworthy AI. At its essence, the framework aims to tackle difficult questions such as legality, safety, and human control, right from the start of a system’s development and through its lifecycle.

The framework is modular, incorporating internal ethical assessments and formalised reviews based on the risk level of the technology. DRDO’s goal here is to entrench responsibility into every stage of AI development and not treat it as an afterthought.

The ETAI framework stands on five key pillars: Reliability, safety, transparency, fairness, and privacy:

Reliability: System performance needs to be accurate, even in chaotic or unfamiliar battlefield scenarios.

Safety: Guardrails against unintended consequences, particularly with autonomous machines.

Transparency: Human commanders should have the ability trace and understand decision chains.

Fairness: Algorithmic biases should not influence critical decisions.

Privacy: Protects sensitive data from unauthorised access.

The more autonomous and impactful an AI system is, the more rigorous its review. For example, a chatbot used for internal training would not require the same oversight as a drone with live targeting capabilities.

In 2021, NITI Aayog also laid out a framework that aligns AI development with India’s constitutional ethos, in particular Article 14 (Right to equality) and 21 (Right to life and personal liberty). The seven principles it emphasised are:

* Safety and reliability

* Equality

* Inclusivity and non-discrimination

* Privacy and security

* Transparency

* Accountability

* Reinforcement of positive human values

Although originally aimed at civilian sectors like health care, these principles are just as relevant for defence. After all, the ethical implications of facial recognition, predictive analytics, and autonomous systems remain just as relevant to military use.

While the policy framed by DRDO and Niti Aayog appears to be robust and well—thought out, turning paper policy into practice is where the real challenge lies. Since these principles lack legal enforceability, co-ordination between agencies and ethical AI frameworks remains limited. As a result, many projects operate in silos, making comprehensive ethical oversight a challenge.

More recently (in 2024), MeitY launched the IndiaAI Mission with over ₹10,000 crore in funding. This is meant to support researchers and companies who need heavy compute power for AI training. A big part of this initiative is (Artificial Intelligence Research, Analytics, and Knowledge Assimilation Platform (AIRAWAT), a public compute cloud designed to support India’s AI needs, especially for startups and public agencies. In practice, AIRAWAT provides a cloud platform where developers can access massive computational power for training AI models. This can be directly beneficial for defence applications such as simulation training, image intelligence analysis, or building large language models for military use, which require supercomputing resources.

While the ministry of defence, MeitY, and NITI Aayog are all making strides, their efforts are largely siloed. India needs a unified platform to align goals, streamline operations, and ensure ethical consistency.

Key recommendations include:

Setting up of a Defence AI Regulatory Authority with clear legal powers

Making ethical and risk-based assessments mandatory in all defence AI procurements

Expand AIRAWAT-like infrastructure specifically for secure military testing environments

Build structured collaboration between DRDO, MeitY, and NITI Aayog

It is clear that India has laid the foundation for responsible AI use in defence. It now needs to bring it all together. As countries in the Global South look for models that balance innovation with ethics, India has an opportunity to lead—not just by building powerful tools, but by using them wisely and lawfully.

This article is authored by Zain Pandit, partner and Aashna Nahar, associate, JSA Advocates and Solicitors.