India should declare that AI will not be used to autonomously launch nuclear weapons

LT GENERAL PRAKASH MENON

Technological advancements like AI, quantum computing, nanotechnology, robotics have made it imperative to modernise the nuclear arsenal.

Twenty-five years have elapsed since India carried out six nuclear tests and announced its ambitions to be a nuclear power.

Pakistan reacted with seven tests. Worldwide condemnation followed. The five permanent members of the United Nations Security Council enforced sanctions. A year later, the Kargil conflict, which was limited in terms of geography and quantum of force, was fought under the nuclear shadow, even though at that time both nations hardly had any operational nuclear weapons capability. Yet it played a role, and to some unknown extent contained the conflict.

In the quarter of the century that has gone by, the nuclear shadow has grown in size and now covers the dyads of India-China and India-Pakistan. The next decade could possibly witness an intensified shift to the India-China dyad, a process that is already underway and is connected to the larger global geopolitical confrontation between the West and the China-Russian combine.

The growth of China’s nuclear capabilities is shaped to counter the capabilities of the USA, but it also influences India’s capabilities which then drives Pakistan’s nuclear capability that has from the very beginning benefited from Chinese assistance. In fact, declassified CIA papers revealed that China provided Pakistan with the design of the nuclear weapon that was tested in Lop Nor, its weapons testing site. The Americans also believe that the test was witnessed by a senior military officer from Pakistan. Such clandestine assistance has continued and can be expected to endure.

Technological challenges

The seemingly intractable territorial disputes between China and India, and India and Pakistan are also overshadowed by the growing global geopolitical tensions which are accompanied by technological developments in artificial intelligence, quantum computing, cyber technology, nanotechnology, robotics, hypersonics and additive manufacturing. The list of technologies is certainly expected to grow. The point to note is that all or most of these technologies possess the potential to impact nuclear weapons and their delivery, as well as surveillance, command and control systems which are the major components of the nuclear arsenal.

All the above technologies could play a major role in the survivability of the nuclear arsenal which is a key element of India’s nuclear doctrine that is based on No First Use. These technologies are also in the dual-use category, spanning civilian and military functions. In addition, they can improve both offensive and defensive capabilities. The nuclear arms race is powered by the reaction between offensive and defensive capabilities. The ongoing tussle between hypersonic missiles and anti-missile capability is an apt illustration. Understandably, technology will continue to chase its own tail.

The Indian challenge is therefore structured within its abilities to develop the technologies and incorporate them in the modernisation of its nuclear arsenal. All this while preserving the prescriptions of its nuclear doctrine that views nuclear weapons as having only political and not military utility. Meeting this challenge will not be easy as it calls for financial resources, civil-military fusion, indigenous capacity and technological cooperation with friendly countries.

Survivability of Command and Control System

The emphasis of the Indian Nuclear Doctrine on civilian control, calls for a robust Command and Control System that must be able to react after being subjected to a nuclear strike. Such a strike could be preceded by a cyber attack that is aimed to paralyse the communications channels.  Continuous strengthening of cyber defence related to nuclear command and control infrastructure must be privileged. Artificial intelligence and quantum computing in particular could also play a major role.

The physical strike can also be focused on the leadership elements that are expected to operate during a crisis from platforms that could be located either in or a combination of underground, undersea, air or ground mobile platforms. All these locations have their practical difficulties to create. They have also their strengths and weaknesses, in terms of concealment, vulnerability and communication.

With the increase in surveillance, accuracy, speed and range of weapons, redundancy of nuclear command locations and communications channels is called for. Normally, it is presumed that for redundancy, one must have at least two political leadership groups. With technological progress, there is a case for examining the necessity of having three groups. What could be ideal is to have two groups that are based on underground facilities and one group that is airborne. If only two groups are practical and preferred, one can be underground and the other airborne.

Costs cannot be a major roadblock for an airborne operational command post due to the payoffs in terms of survivability and strengthening of communications.

Quest for survivability 

India’s nuclear arsenal is based on a triad of air, sea and land domains that is also dependent on satellites in outer space. All these have the potential to increase our survivability due to mobility and the vastness of India’s continental and maritime space.

The air leg is relatively more vulnerable because of the dependency on fixed assets and the difficulty of concealment. The sea leg based on the nuclear-powered ballistic missile submarine, SSBN, is considered the least vulnerable but its communication links with civilian and military leaders could be increasingly vulnerable due to cyber attacks on the ground stations and satellites. The land leg can be distributed across the massive continental space and also have mobility. Overall, it would be prudent to prioritise our capabilities in the order of SSBN, land-based ballistic and airborne nuclear vectors. But the challenge would be making distribution decisions within the triad, while preserving the principle of Credible Minimum Deterrence and avoiding pursuing numbers in the quest for survivability.

One of the key assumptions must be of eschewing the notion of a surprise disarming first strike, often described as a ‘bolt from the blue’ by either China or Pakistan. This is a military imagination that supposes that either China or Pakistan, would, without a major ongoing crisis, decide to use its nuclear weapons to neutralise all of our nuclear capability.

Apart from the military difficulty of carrying out such a strike, it would be tantamount to the entire subcontinent and even China becoming victims to the ravages of climatic changes induced by nuclear weapon explosions – the dust, smoke that blocks out the sun and radioactive fallout that would have long-term effects on bio-systems and impact water and food supply. Scientific simulations have pointed to such a ‘nuclear winter’. But this is an inconvenient probability that has been ignored by nuclear powers especially those that retain the threat of ‘First Use’.

Low-yield weapons have been touted as an alternative to prevent an escalation spiral. This is a theoretical proposition that is unlikely to be realised, as once nuclear weapons are used between nuclear powers, the reaction is unpredictable and would be subjected to the vagaries triggered by the prevailing ambience of fear, danger and uncertainty that has to grapple with miscommunication, misinformation and misperception.

India should declare that its artificial intelligence capabilities will not be used for autonomously launching nuclear weapons.  Such an announcement has already been officially made by the USA. This could be another realm for India-USA cooperation.

A turbulent geopolitical decade is in the offing. In the context of the India-China-Pakistan equations, while technology will play a major role, India must not fall prey to human wisdom being overwhelmed by the promises of machines that knock on the doors of its nuclear doctrine, which must endure.

Lt Gen (Dr) Prakash Menon (retd) is Director, Strategic Studies Programme, Takshashila Institution; former military adviser, National Security Council Secretariat. He tweets @prakashmenon51. Views are personal.

(Edited by Theres Sudeep)