- Lt Gen Raj Shukla called AI-powered drones a historic shift in national security at NDTV Defence Summit 2026
- Ukraine doubled drone operations using algorithmic warfare, showcasing the scale of AI in combat, he noted
- India must adapt structurally and culturally to AI drone warfare to meet security challenges, he further said
India must see the use of AI-powered drones in combat as a "national security epoch", or it will lag in modern warfare, Lt General Raj Shukla (retd) stressed on Sunday, speaking at the NDTV Defence Summit 2026 in New Delhi.
"This is [one of] those once-in-three-centuries moments... like gunpowder, mechanisation, tanks, airpower, those kinds of things. And it's very important to acknowledge this because, unless we do that, we will not step up," the former Army Commander said in a panel discussion moderated by NDTV's Vishnu Som.
He explained with Ukraine as an example: "Till last year, the locus (control) of drone operations was with some very skilled remote operators. Because of the limitation of operators, Ukraine could launch 10,000 drone operations a day. This year, because of the transition to algorithmic warfare, they are launching about 20,000 operations a day. Look at the scale."
"And it's across theaters," he added. Ukraine has 8 million armed drones, according to Lt Gen Shukla. "Russia, having learnt from Ukraine, has the Geran program, 10 million [drones]. The United States is faltering at about 22,000, though now they have the Drone Dominance Initiative [at a budget of] $70 billion. China has 1 billion armed drones. That should wake us up -- 1 billion," he added.
"The authorisation to strike has moved away from humans to computer vision. That's already happened. So it is no longer human in the loop; it is on the loop," he added.
He said India would need to adapt, "make structural and cultural corrections", which he termed "bigger than the China challenge", referring to tensions with the eastern neighbour. "Unless we get this (AI) adaptation right, we will not be able to measure up to other challenges, like China and all," he stressed.
Forecasting, he further said, "You will have lethal, precise drone swarms, autonomous targeting, constantly getting better because of recursive learning. I am confident that by next year it will be humans out of the loop."
He cited a recent interview by Admiral Samuel Paparo, Commander of the United States Pacific Command, about how the US could potentially deter a Chinese amphibious assault by turning a defined area into "a hellscape of drones, missiles, autonomous systems, powered by AI".
"All this is being powered by AI software: Delta, Maven, Claude. If we are not talking about that in India, then we seem to be behind... I don't think we have quite grasped the stakes," said Lt Gen Shukla, who is currently a member of the Union Public Service Commission.
The question of how soon India could have fully AI-enabled drones was put to another panellist, Sameer Joshi, CEO of NewSpace Research and Technologies.
Joshi, a retired Indian Air Force fighter pilot, explained, "Largely autonomous drones powered by AI software and automatic target recognition are able to close the kill chain in the shortest possible time. They can recognise a target and go after it. But he added caution: "The Ukrainians, talking of Hornet (drone system) -- which is interestingly made by a company which is supported by Eric Schmidt, the former CEO of Google -- say that 'the drone proposes and the human on the loop disposes'. I ask: in a dense electronic warfare environment, who is going to dispose the decision?" However, he said autonomous drones are "here to stay". "It is on the verge, on the bounds of autonomy, trust, ethics, and how you want to use it."
India is likely to absorb fully autonomous drones with automatic target recognition from next year, he said. "But India, like many other countries, is also a signatory to a pact where it conveys that if we use drones ethically, the requirement largely will centre around the human in the loop," he said.
But is India ready to hand over control to autonomous units?
This was a question to Suresh Kamate, General Manager, Aerospace and Defence Business, dSPACE India.
"That's a fundamental question of trustworthy autonomy," Kamate said. "We don't just hope AI makes the decision, but we prove it, and under the conditions which are far more dangerous than what this AI system sees in the battlefield," he said.
"At dSPACE, we do this through a digital engineering chain... [It] creates a virtual battlefield and helps us to refine the decision-making engine... in real time, before it moves from the virtual battlefield to a real-world scenario," he said. "We provide a platform where the decision-making algorithms get tested before putting them into the real hardware," he further explained.
