Ever since Google DeepMind’s AlphaGo program defeated South Korea’s top professional Lee Sedol in 2016 in the popular board game Go, artificial intelligence (AI) technologies, including machine learning and deep learning, have seized the imagination of people across the globe. While the impact of AI is already being felt in many areas, such as speech recognition in digital assistants like Siri and Cortana, and consumer behaviour prediction by Amazon and Google, it is future AI systems that are creating all the excitement. The most prominent amongst these is self-driving cars, with the year 2020 being targeted by market leaders for manufacturing cars capable of driving themselves without any human intervention.
There appears to be general agreement on the positive benefits which AI can bring to society. At the same time, there is also an underlying fear grounded in the belief that AI systems would one day exceed human intelligence and capabilities, a line of thinking which leads to many doomsday scenarios. Stephen Hawking, Bill Gates, Elon Musk and others have all expressed serious concern about the uncontrolled development of AI applications, stating repeatedly that AI could pose the greatest existential threat to humanity.
Of immediate concern, however, is the use of AI in military applications, specifically those termed as Lethal Autonomous Weapon Systems (LAWs). LAWs, according to a commonly accepted definition, are weapon systems that “once activated, can select and engage targets without further human intervention. The prospects of this military application has given rise to the Campaign to Stop Killer Robots, a global coalition of 64 non-government organisations (NGOs) launched in April 2013 under the aegis of Human Rights Watch with the aim of pre-emptively banning fully autonomous lethal weapons. This Campaign, amongst others, advocates the view that retaining human control over the use of force is a moral imperative and essential for promoting compliance with international law and ensure accountability.
An informal group of experts from a large number of countries has been debating the issue of LAWs for three years now at the United Nations Office of Disarmament Affairs (UNODA) forum known as Convention on Certain Conventional Weapons (CCW). In December 2016, countries agreed to formalise these deliberations. As a result, a Group of Governmental Experts (GGE) has been established, the first of which was held from 13 to 17 November 2017, chaired by Ambassador Amandeep Gill of India. Approximately 90 countries along with many other agencies participated in the meeting. Some of the conclusions arrived at during the meeting were:
It was also agreed that a ten-day meeting should be scheduled in 2018.
The primary argument put forth by advocacy groups calling for a ban is that weapon systems that have autonomy in the critical functions of ‘select and engage’ would be in violation of International Humanitarian Law (IHL), and specifically its principles of distinction and proportionality. While the former principle requires weapon systems to be able to reliably distinguish between combatants and civilians, the latter requires value judgement to be used before applying military force. According to this argument, LAWs will never be able to live up to these requirements. In addition, there is also the consideration of what is known as the Martens Clause, wherein it is contended that delegating to machines the decision power of ‘life and death’ over humans would be “against the principles of humanity and the dictates of public conscience.”
There is an equally vocal body of opinion which states that the development and deployment of LAWs would not be illegal, and in fact would lead to the saving of human lives. This is because without the driving motivation for self-preservation, LAWs may be used in a self-sacrificing manner, saving human lives in the process. Moreover, they can be designed without emotions that normally cloud human judgment during battle leading to unnecessary loss of lives. An argument is also put forth that autonomous weapons would have a wide range of uses in scenarios where civilian presence would be minimal or non-existent, such as tank or naval warfare, and that the question of legality depends on how these weapons are used, and not on their development or existence.
Some of the well-known autonomous defensive weaponry already in use today are missile defence systems such as the Iron Dome of Israel and the Phalanx Close-In Weapon System used by the US Navy. Fire-and-forget systems, such as the Brimstone missile system of the United Kingdom (UK) and the Harpy Air Defense Suppression System of Israel, also function in an autonomous manner. Another oft quoted example of existing autonomous weapon systems is the SGR-A1, a sentry robot deployed by South Korea in the Demilitarized Zone with its northern adversary.
The relevance of the ongoing debate on LAWs in the context of the Indian military landscape cannot be over-emphasized, as there are many scenarios where these can be deployed to advantage. Autonomous systems designed to disarm improvised explosive devices (IEDs) are already in use by Indian forces, although these are non-lethal and defensive in nature. Future possible applications include AI-enabled drone swarms to boost surveillance capabilities; robot sentries along the borders to check infiltration by terrorists; autonomous armed UAVs for use in conventional as well as sub-conventional scenarios, and so on. In general, saving own soldiers from the lethality of war would yield rich dividends to any military force, especially in conventional conflicts. Faced with the prospect of a two and a half front war, the development of LAWs by India assumes strategic significance.
The United States has put AI at the centre of its quest to maintain its military dominance. As a part of its Third Offset Strategy announced in 2014, the Pentagon has reportedly dedicated US $18 billion for its Future Years Defense Program, a substantial portion of which has been allocated for robotics, autonomous systems and human-machine collaboration. Chinese military leaders and strategists believe that the character of warfare is fundamentally changing due to unmanned platforms and autonomous systems and have labelled AI research as a national priority, thus giving it a huge impetus. Russian president Vladimir Putin has recently predicted that whichever country leads the way in AI research will come to dominate global affairs. Without doubt, Russia is pursuing the development of LAWs in earnest, in order to keep pace with the US and China in this new arms race. The UK, France and Israel, amongst others, are expected to be significant players in this contentious new field.
In the context of India’s defence, presently there appears to be a void in terms of doctrines and perspective plans when it comes to the exploitation of AI/Robotics technologies. Occasional interactions by the defence establishment with the DRDO’s Centre for AI and Robotics (CAIR) and other agencies are inadequate to spur the latter into producing timely and meaningful results. Given its track record, DRDO is unlikely to be successful in developing complex lethal autonomous systems anytime soon. It is also worth noting that world-wide, R&D in these technologies is being driven by the private commercial sector rather than the defence industry. Unfortunately, the Indian equivalents of Baidu, Amazon, Google and Microsoft are yet to rise to the occasion, despite the strengths of our IT industry. Clearly, much more needs to be done.
After decades of false starts, AI/Robotics technologies today appear to be at an inflection point, making rapid advancements which are considered significant enough to usher in a new revolution in military affairs (RMA). Notwithstanding the world-wide concern about the development of LAWs from the legal and ethical points of view, it is increasingly clear that, irrespective of any conventions that may be adopted by the UN, R&D by major countries is likely to proceed unhindered. Given India’s security landscape, perhaps there is a need to adopt a radically different approach for facilitating the development of LAWs. As with any transformation, this is no easy task. Only a determined effort, with specialists on board and due impetus being given from the apex level, is likely to yield the desired results.