Kaustubh Kumar & Ashutosh Anand*
“Life on the planet began in the ocean. The ocean sustains it. Undoubtedly perils of the ocean will have an effect on the lives on earth.”
Prabhakaran Paleri, PTM, TM,
Director-General, Indian Coast Guard (Retd.)[i]
Artificial Intelligence (AI) has initiated a revolutionary phase in human life. It has the potential to transform every sphere of humanity, such as the use of AI in environmental conservation, legal system to maintain law and order, healthcare sector, space industry, and defence sector. The recent killing of Iranian nuclear scientist Mohsen Fakhrizadeh through a satellite-controlled weapon platform, which has a remotely controlled machine gun mounted on a pickup, was the first instance where a machine gun with the help of AI was used for attacking an individual; however, no collateral damage was caused.[ii]
Another recent example of the use of AI in the defence sector can be the AI-powered ‘Iron Dome’ anti-missile system, which is reportedly 90% effective in automatically identifying, countering, and destroying rockets, shields and mortars, as well as aircraft, helicopters, and unmanned aerial vehicles.[iii] This defence system destroyed approximately all of the rockets fired by Hamas on Israel.[iv] Moreover, the concept of ‘Slaughterbots’[v] or ‘Killer robots’,[vi]AI-powered remote-controlled attack drones, are also in the news for the past few years, giving us an apt example to demonstrate how AI might help augment future security and surveillance mechanisms of defence forces on land and air.[vii]
However, the AI might also play a crucial role in defending the maritime borders of a state by enhancing the capability of Naval forces. The creation of fully autonomous systems with lethal capabilities bolsters the claims that AI-enabled battlefields are the future. India recently completed the second phase of an AI-enabled system – Trigun, which aims to enhance maritime battlespace transparency with better network-centric warfare tools riding on high-speed data communication systems and integration.[viii] Although these technologies are developing day by day at a higher pace, the question of ethics looms.
Therefore, this blog analyses why it is a need of the hour to focus on the question of whether AI is necessary for maritime security and what are threats that it poses to human life. In this blog, an attempt has been made to represent how AI can help in mitigating the issues faced while dealing with maritime security.
Need of AI-Enabled Maritime Security
The maritime sector is considered the cornerstone of international trade and commerce as shipping is the most preferred way to transport goods from one region to another. According to the United Nations Conference on Trade and Development (UNCTAD), the sea route witnesses around 70 percent of trade by worth and 80 percent by volume.[ix] However, shipping is vulnerable to numerous challenges, such as ship hijacking, piracy, and armed robbery, maritime terrorism, illegal fishing, drugs, and human trafficking, to name a few.
Due to the high dependency of world trade on the maritime sector, even a small change in geopolitics has a drastic impact. For instance, friction between Iran and the United States rose in May 2019, when the USA sent more military assets to the region of Persian Gulf in response to intelligence reports of an alleged ‘campaign’ by Iran and its ‘proxies’ to endanger US soldiers and oil shipments via the Strait of Hormuz.[x] In the contemporary world, we have seen intense geopolitical clashes. With the rapid changes in technology combined with the debilitating geopolitical scenario, the need for innovation in maritime security increases.
Furthermore, recently, in 2020, after five days of fierce clashes between the insurgents and Mozambican security forces, the Islamist insurgents captured a strategic port of Mocimboa da Praia in the restive province of Cabo Delgado in northern Mozambique. The port was at a strategic site where gas companies are operating as well as different multinational companies have their huge investments in the area.[xi] However, if the Mozambican navy had such modern warfare capabilities laced with AI, then the scenario would have been different. Thus, such instances of maritime terrorism that not only pose a threat to the shores of a country but to the sea vessels also shall get mitigated with the use of AI.
How Can AI Help in Managing Maritime Security?
In simple terms, AI can be defined as information systems capable of using computer algorithms or rule-sets to do things. A computer or machine can learn from examples and experiences and impersonate them in reality.[xii] It can help in resolving the issues faced in securing maritime trade.
Oceanic regions are typically unmapped and challenging to navigate. The employment of AI-based systems to track, compute, identify, chart, and execute the appropriate actions for a vessel enhances existing maritime capabilities.[xiii] AI can also bolster the ability of intelligence gathering, surveillance and reconnaissance. Furthermore, the data retrieved through these methods can be used to identify and engage enemy targets.AI can also help by sending automatic emergency signals to the nearest naval ship or base to rescue any passenger ship or vessel carrying human life. AI can also provide real-time data about the location of the ship, thus safeguarding the goods and the lives of passengers and crew-members aboard.
AI can also help in mitigating the aerial, underwater, and surface threats posed by enemy warships, submarines, etc., by enhancing the capability of naval forces. As supersonic fighter jets have become a new trend, naval ships with enhanced AI capability can monitor and precisely predict their flight trajectory, which might help a country in dealing with the alien threat posed by such jets. Quick data sharing and automated handling of warships with the help of AI can augment the operational capability by rapidly deploying fighting units of warships in any fast-moving war scenario. AI-enabled unmanned underwater vehicles can play a significant role in combat operations. Thus, the infusion of AI with naval forces shall help in protecting the maritime borders covering another crucial aspect of maritime security.
The AI-enabled maritime security systems have one thing in common: their capacity to continually learn from the inputs they get from their surroundings. Deep learning algorithms incorporated in their systems enable them to learn from a variety of situations over time, particularly through assessing human input. Developers of artificial intelligence have realised that depending entirely on human input is insufficient. As a result, these systems are increasingly relying on ‘machine learning’ to speed up the learning process.[xiv] “Machine learning is a method to software development in which systems learn tasks and improve their performance via experience,” according to a report.[xv] Thus, this way, AI has enough potential to resolve the issues challenging maritime security.
Challenges in AI-Enabled Security Mechanisms’ Implementation
Nevertheless, as these technologies advance, ethical and legal concerns about the employment of autonomously competent deadly AI-enabled weapons are becoming more prominent. The level of autonomy that the AI security systems are expected to acquire leaves the question of whether the world will concede to such a level. The Geneva Convention’s Article 36 (Additional Protocol I)[xvi] mandates that states conduct a legal assessment of new weapons to make sure that they comply with international law. AI-assisted systems with targeting capabilities will need to be able to differentiate between military and civilian assets on a continual basis.
As more and more AI-assisted security systems enter the domain of maritime security in the future, the likeliness of resistance is also expected to increase. The most significant of them is institutional resistance from traditional naval personnel, who will resist any attempt to replace them. Moreover, a new training curriculum will be required to generate technologically literate and AI-calibrated employees, as well as a testing and validation framework to test these new technologies.[xvii]Operational concepts and tactics, in particular, must be redesigned to accommodate multi-vehicle control and AI-enabled maritime regime.
The vulnerability of such systems to countermeasures is another obstacle to their development and implementation. Many AI-assisted systems are still in the design phase, and it is safe to assume that enemies will invest in countermeasures in parallel with their development. For example, acoustic stealth and acoustic quieting technologies are currently being incorporated into submarines, and future advancements in these technologies might enable the development of an AI-based vessel’s acoustic identification system by its adversaries.[xviii]
Moreover, the autonomous ships would be more prone to hijacking and piracy as there would be no crew on-board that is likely to assist or avert any piracy attack, thereby making the work of pirates easier. The autonomous ships would be operated through navigational sensors in contact with an AI-control room monitoring the ship. As soon as pirates board the ship, they can destroy such sensors making the vessel inoperative by disturbing the networking between the vessel and control room. It is rationalised that due to the non-existence of crews on-board, instances of kidnapping and getting ransom would plunge. However, the chances of ship hijacking and then demand of ransom to release the ship or further use of the vessel for smuggling of drugs and weapons is likely to increase.[xix]
Further, the AI-enabled defence mechanisms also have a significant adverse impact on marine life. The underwater platforms primarily rely on SONAR transmissions to transmit or gather any information. Such transmissions can cause potential damage to corals as well as to the sensory organs of marine species such as whales, sharks, dolphins, and other migratory species. These marine species are already classified under ‘critically endangered,’ ‘endangered,’ and ‘extinct’ categories in IUNC Red List.[xx] Usage of such transmissions might result in the disorientation of their body or even death.
The undersea topography, such as trenches, rocks, mounds, ridges, etc., and underwater activities, such as laying of optical fibre cables, oil and gas pipelines might pose a threat like interception and deviation of signals sent by any AI-enabled unmanned underwater vehicle or base.[xxi] Thus, such issues need to be looked upon before the implementation of such a mechanism for its smooth functioning and the sustenance of the environment.
“Computers will overtake humans with AI at some within the next 100 years. When that happens, we need to make sure the computers have goals aligned with ours.”
~ Stephen Hawking (Zeitgeist Conference, 2015)[xxii]
AI can be proved as a beneficial tool in the alleviation of the issues related to Maritime Security during recent times. The technological advancement, which is certain to happen, should be rendered in a cautious manner by creating enough sophisticated machines that, except an engineer knowing its intricacies, cannot disarray the system. Further, in the case of autonomous ships, the algorithms programmed must be able to make prompt decisions and determine their actions. The sensors and system with algorithms containing navigational protocols must be placed deep inside the vessel, which makes it more or less impossible for an individual to interfere with them. Non-lethal weapons can also be used in autonomous ships to prevent any act of hijacking.
There are lacunae and inadequacies in AI-related laws and international regulations at present. For instance, flag states are obligated by certain duties under the United Nations Convention on the Law of the Sea (UNCLOS). These duties comprise maintaining that all ships are under the command of a master and officers who have the necessary qualifications. However, the international regulations become uncertain in the absence of master and crew on an AI-enabled vessel.[xxiii] Therefore, it will be pertinent to fill the gaps in international law in future.
At present, AI applications still depend on human operators. The attacks are conducted by AI-enabled systems, but the ultimate decision of target engagement rests on humans. Therefore, it is imperative to enhance AI systems to understand human decision-making, and the chances of ‘attacks by mistakes’ can be reduced.
Most significantly, unmanned vessels engaged in long-term floating missions must be powered by renewable energy sources to assure that their missions do not degrade the marine ecosystem.
The essay also delineated the challenges and threats that AI-enabled systems may pose. The maritime ecosystem is gigantic and complicated. Securing the ecosystem from being attacked can be done easily with the assistance of AI. Thus, to pave a path for success, the challenges should be taken care of, and concrete steps must be taken to prevent them.
About the Authors
Kaustubh Kumar is a law student at the National University of Study and Research in Law, Ranchi, India. He has a keen interest in legal research and writing. His area of interest lies in Cyber Law and Public policy issues.
Ashutosh Anand is a law student at the National University of Study and Research in Law, Ranchi, India. He has a keen interest in legal research and writing. He gets fascinated by constitutional and international law and loves debating on a varied range of topics.
[i]Prabhakaran Paleri, Integrated Maritime Security: Governing the Ghost Protocol 47 (1 ed. Vij Books India Pvt Ltd 2014).
[ii] Huma Siddiqui, Future warfare: Is Indian Army ready for the use of Artificial Intelligence and Smart Technologies?,Financial Express(Dec. 8, 2020, 3:28 PM), https://www.financialexpress.com/defence/futre-warfare-is-indian-army-ready-for-the-use-of-artificial-intelligence-and-smart-technologies/2145585/.
[iii]Debolina Biswas, Israel’s Iron Dome Puts AI At The Forefront Of Modern Warfare, analyticsindiamag.com (May 19, 2021), https://analyticsindiamag.com/israels-iron-dome-puts-ai-at-the-forefront-of-modern-warfare/
[iv]Iron Dome defence system explained: Why Hamas rockets fail to hit targets in Israel?,The New Indian Express (May 12, 2021, 08:24 PM), https://www.newindianexpress.com/galleries/world/2021/may/12/iron-dome-defence-system-explained-why-hamas-rockets-fail-to-hit-targets-in-israel-103135–1.html
[v]Stuart Russel et al, Why You Should Fear “Slaughterbots”—A Response, Spectrum.IEEE.Org (Jan. 23, 2018), https://spectrum.ieee.org/why-you-should-fear-slaughterbots-a-response.
[vi]Ingvild Bode, The threat of ‘killer robots’ is real and closer than you might think, The Conversation (Oct. 16, 2020, 12:01 AM), https://theconversation.com/the-threat-of-killer-robots-is-real-and-closer-than-you-might-think-147210.
[vii] Paul Scharre, Why You Shouldn’t Fear “Slaughterbots”, Spectrum.IEEE.Org (Dec. 22, 2017, 2:45 PM), https://spectrum.ieee.org/automaton/robotics/military-robots/why-you-shouldnt-fear-slaughterbots.
[viii] Mayank Singh, New system with AI to boost maritime security, New Indian Express (Feb. 9, 2020, 10:47 AM), https://www.newindianexpress.com/thesundaystandard/2020/feb/09/new-system-with-ai-to-boost-maritime-security-2100873.html.
[ix]Review of Maritime Transport 2018, UNCTAD, https://unctad.org/webflyer/review-maritime-transport-2018 (last visited May 27, 2021).
[x] Megan Specia et al., The Tension Between America and Iran, Explained,NY Times (May 16, 2019), nytimes.com/2019/05/16/world/middleeast/iran-tensions-explainer.html
[xi] Andre Baptista, SirwanKajjo, Islamist Insurgents Capture Strategic Port in Northern Mozambique, voanews.com (Aug. 13, 2020, 01:27 PM), https://www.voanews.com/africa/islamist-insurgents-capture-strategic-port-northern-mozambique
[xii] IBM Cloud Education, Artificial Intelligence (AI), IBM.Com (Jun. 3, 2020), https://www.ibm.com/in-en/cloud/learn/what-is-artificial-intelligence
[xiii]The Weaponization of Increasingly Autonomous Technologies in the Maritime Environment: Testing the Waters, United Nations Institute for Disarmament Research, 2015, http://www.unidir.ch/files/publications/pdfs/testing-the-waters-en-634.pdf
[xiv]MaaikeVerbruggen and Vincent Boulanin, Mapping the Development of Autonomy in Weapon Systems, Stockholm International Peace Research Institute (Nov. 2017), https://www.sipri.org/publications/2017/other-publications/mapping-development-autonomy-weapon-systems
[xvi]Additional Protocol I to the Geneva Conventions, United Nations, https://www.un.org/en/genocideprevention/documents/atrocity-crimes/Doc.34_AP-I-EN.pdf (last visited May 27, 2021).
[xvii]Verbruggen, supra note 10.
[xviii] Kris Osborn, Is the US Navy Getting Ready to Build the Most ‘Stealth’ Submarine Ever?,The National Interest (Sep. 1, 2016), http://nationalinterest.org/blog/the-buzz/the-us-navy-getting-ready-build-the-most-stealth-submarine-17549
[xix]Thibaut Eude, Examining Autonomous Ships’ Vulnerability to Piracy, The Maritime Executive, (Sep. 14, 2019, 09:53 PM), https://www.maritime-executive.com/blog/examining-autonomous-ships-vulnerability-to-piracy.
[xxi] Vijay Sakhuja, The Challenges of AI-enabled Underwater Platforms, Indian Defence Review (Aug. 08, 2018), http://www.indiandefencereview.com/spotlights/the-challenges-of-ai-enabled-underwater-platforms/
[xxii] Brian Albert, Stephen Hawking: Computers Will Overtake Humans in 100 Years, IGN Brasil (Nov. 13, 2015 09:04 PM), https://br.ign.com/news/3910/stephen-hawking-computers-will-overtake-humans-in-100-years
[xxiii] Nic Gardner, How will AI impact Maritime Law?,Thetius, https://thetius.com/how-will-ai-impact-maritime-law/ (last visited Aug. 9, 2021).