AI in the military: India’s path to ethical and strategic leadership
This article is authored by Zain Pandit, partner and Aashna Nahar, associate, JSA Advocates and Solicitors.
In earlier articles, we traced India’s path through the building blocks of military AI: From institutional frameworks, to ethical and strategic imperatives, and comparative global approaches. This article looks ahead at how India can reinforce its domestic structures, borrow effectively from global best practices, and emerge as a normative leader for the Global South.
India has already started laying down a framework for responsible military AI. In 2024, the Defence Research and Development Organisation (DRDO) introduced the ETAI framework (Evaluating Trustworthiness in AI) in an effort to ensure that the country’s use of AI in warfare is not only cutting-edge but also principled. ETAI is built on five fundamental pillars: reliability, safety, transparency, fairness, and privacy. While these principles may appear abstract, their objective is clear, which is to ensure that military personnel, decision-makers, and the public can place trust in the technology deployed on their behalf.
To reinforce the ETAI framework, India created new decision-making bodies such as the Defence AI Council (DAIC) and the Defence AI Project Agency (DAIPA). These were set up to guide projects, coordinate research, and give direction to an area of technology that is advancing at breakneck speed. The institutional design reflects clarity of purpose and intent. But so far, these bodies are conveners, not enforcers. They bring stakeholders together but cannot compel compliance. The gap India must now close is the one between principle and practice.
India can take valuable cues from the United States Department of Defense (DoD). The DoD’s Responsible AI principles—responsibility, equity, traceability, reliability, and governability— are translated into clear operational practices. Oversight rests with the Chief Digital and AI Office, which ensures accountability runs through the chain of command. Ethical principles are backed by impact assessments and trustworthiness evaluations, producing systems that are not only tested but trusted.
Trust, in this context, comes from more than just rigorous testing. It depends on governance structures that can be relied upon and a workforce that is properly trained and educated. By investing in people as much as in technology, the DoD has made AI something its personnel can understand, work with, and rely on—not a black box to be feared.
The US model shows how to overcome the last mile problem of ethics: Turning values into practice. For India, this is the crucial step ahead. Mandating traceability in every deployed system, making human override a non-negotiable, and—most importantly—creating a statutory Defence AI Regulatory Authority with powers to certify, investigate, and enforce compliance would give substance to the ethical frameworks already discussed. Without enforcement, India risks strong ideas but weak execution.
The way forward lies in ambition matched with reform. An independent authority with regulatory teeth can ensure that ethical safeguards are not just discussion points. Its mandate should include certification, investigation, and enforcement. Ethical and risk-impact assessments should be mandatory for all defence AI procurement, drawing on global best practices. At the same time, India must invest in infrastructure for stress-testing AI: The AIRAWAT initiative—already providing cloud power and datasets for civilian AI—could be adapted into a defence-specific platform to simulate battlefield pressures and test AI systems under extreme conditions. Just as important is human capacity. India needs to train and upskill its defence and technical workforce to use AI responsibly, combining technical skill with ethical awareness. Only then can AI become a trusted partner in national security.
Domestic reform is only part of the story. India has a bigger opportunity: to lead globally, especially for the Global South. Many developing countries want AI in defence but lack resources to design governance from scratch. They need templates that combine innovation with restraint. India is well placed to provide this. Sharing frameworks like ETAI, offering training programmes, and extending advisory support could make India a partner of choice. Initiatives like the Defence India Startup Challenge (DISC) already nurture technologies that could be adapted for international collaborations. At forums like the UN GGE, G20, and BRICS, India can push for inclusive governance that reflects not just western concerns but also the realities of resource-constrained states. That means advocating equitable access to AI benefits, feasible accountability standards, and human control as a universal principle.
This article is authored by Zain Pandit, partner and Aashna Nahar, associate, JSA Advocates and Solicitors.
