Author: Shashank Yadav

  • Social automation and APT attributions in National Cybersecurity

    Social automation and APT attributions in National Cybersecurity

    Advanced Persistent Threats (APTs) are a prime concern in the formulation and implementation of national cybersecurity policies. These threats often also involve complex social engineering tactics which are undergoing a quantitative and qualitative revolution with burgeoning AI capabilities. However, the attribution of these APT activities can be mired with technical and political considerations. We analysed 142 APT groups’ attributions along with their use of social interaction vectors to ascertain the nature of the risk environment and the operational threat landscape of AI and social automation. We discover that close to 80% of APT activities could be chalked up to merely 20% of key nation-state threat actors competing with each other. We further discuss the implications of this political threat distribution for national cybersecurity environments.

    Keywords: cybersecurity, AI Policy, advanced persistent threats, automation, social engineering


    Read full paper here

  • Ghosts in the Machine: The Past, Present, and Future of India’s Cyber Security

    Ghosts in the Machine: The Past, Present, and Future of India’s Cyber Security

    [powerkit_button size=”lg” style=”info” block=”false” url=”https://admin.thepeninsula.org.in/wp-content/uploads/2022/10/Research-Paper-TPF-1.pdf” target=”_blank” nofollow=”false”]
    Download
    [/powerkit_button]

    Introduction

    When the National Cybersecurity Policy was released in 2013, the response from experts was rather underwhelming [1], [2]. A reaction to a string of unpalatable incidents, from Snowden’s revelations [3] and massive compromise of India’s civilian and military infrastructure [4] to the growing international pressure on Indian IT companies to fix their frequent data breaches [5], the 2013 policy was a macro example of weak structures finding refuge in a haphazard post-incident response. The next iteration of the policy is in formulation under the National Cybersecurity Coordinator. However, before we embark upon solving our cyber-physical domain’s future threat environment, it is perhaps wise to look back upon the perilous path that has brought us here.  

    Early History of Electronic Communications in India

    The institutional “cybersecurity thinking” of post-independence Indian government structures can be traced to 1839 when the East India Company’s then Governor-General of India, Lord Dalhousie, had asked a telegraph system to be built in Kolkata, the then capital of the British Raj. By 1851, the British had deployed the first trans-India telegraph line, and by 1854, the first Telegraph Act had been passed. Similar to the 2008 amendment to the IT Act which allowed the government to intercept, monitor and decrypt any information on any computer, the 1860 amendment to the Telegraph Act too granted the British to take over any leased telegraph lines to access any of the telegraphs transmitted. After all, the new wired communication technology of the day had become an unforeseen flashpoint during the 1857 rebellion.

    Post-independence, under the socialist fervour of Nehruvian politics, the government further nationalised all foreign telecommunications companies and continued the British policy of total control over telecommunications under its own civil service structure, which too came pre-packaged from the British.

    Historians note that the telegraph operators working for the British quickly became targets of intrigues and lethal violence during the mutiny [6], somewhat akin to today’s Sysadmins being a top social engineering priority for cyber threat actors [7]. One of the sepoy mutineers of 1857, while on his way to the hangman’s halter, famously cried out at a telegraph line calling it the cursed string that had strangled the Indians [8]. On the other side of affairs, after having successfully suppressed the mutiny, Robert Montgomery famously remarked that the telegraph had just saved India [9]. Within the telegraph system, the problems of information security popped up fairly quickly after its introduction in India. Scholars note that commercial intelligence was frequently peddled in underground Indian markets by government telegraph clerks [10], in what can perhaps be described as one of the first “data breaches” that bureaucrats in India had to deal with. 

    British had formulated different rules for telecommunications in India and England. While they did not have the total monopoly and access rights over all transmissions in Britain, for the purpose of maintaining political control, in India they did [11]. Post-independence, under the socialist fervour of Nehruvian politics, the government further nationalised all foreign telecommunications companies and continued the British policy of total control over telecommunications under its own civil service structure, which too came pre-packaged from the British.

    The Computer and “The System”

    Major reforms are often preceded by major failures. The government imported its first computer in 1955 but did not show any interest in any policy regarding these new machines. That only changed in 1963, when the government under the pressure to reform after a shameful military defeat and the loss of significant territory to China, instituted a Committee on Electronics under Homi Jehangir Bhabha to assess the strategic utilities that computers might provide to the military [12].  

    In 1965, as punitive sanctions for the war with Pakistan, the US cut off India’s supply of all electronics, including computers. This forced the government to set up the Electronics Committee of India which worked alongside the Electronics Corporation of India (ECIL), mandated to build indigenous design and electronic manufacturing capabilities. But their approach was considered highly restrictive and discretionary, which instead of facilitating, further constrained the development of computers, related electronics, and correspondingly useful policies in India [13]. Moreover, no one was even writing commercial software in India, while at the same the demand for export-quality software was rising. The situation was such that ECIL had to publish full-page advertisements for the development of export-quality software [12]. Consequently, in the early 1970s, Mumbai-based Tata Consultancy Services managed to become the first company to export software from India. As the 1970s progressed and India moved into the 1980s, it gradually became clearer to more and more people in the government that their socialist policies were not working [14]. 

    In 1984, the same year when the word ‘Cyberspace’ appeared in a sci-fi novel called Neuromancer, a policy shift towards computing and communications technologies was seen in the newly formed government under Rajiv Gandhi [12]. The new computer policy, shaped largely by N. Sheshagiri who was the Director General of the National Informatics Centre, significantly simplified procedures for private actors and was released within twenty days of the prime minister taking the oath. Owing to this liberalisation, the software industry in India took off and in 1988, 38 leading software companies in India came together to establish the National Association of Software and Service Companies (NASSCOM) with the intent to shape the government’s cyber policy agendas. As we are mostly concerned about cybersecurity, it should be noted that in 1990, it was NASSCOM that carried out probably the first IT security-related public awareness campaign in India which called for reducing software piracy and increasing the lawful use of IT [5].   

    Unfortunately, India’s 1990s were mired by coalition governments and a lack of coherent policy focus. In 1998, when Atal Bihari Vajpayee became the Prime Minister, the cyber policy took the most defining turn with the development of the National IT Policy. The IT Act, thus released in 2000 and amended further in 2008, became the first document explicitly dealing with cybercrime. Interestingly, the spokesman and a key member of the task force behind the national IT policy was Dewang Mehta, the then president of NASSCOM. Nevertheless, while computer network operations had become regular in international affairs [15], there was still no cyber policy framework or doctrine to deal with the risks from sophisticated (and state-backed) APT actors that were residing outside the jurisdiction of Indian authorities. There still is not.  

    In 2008, NASSCOM established the Data Security Council of India (DSCI), which along with its parent body took it upon itself to run cybersecurity awareness campaigns for law enforcement and other public sector organisations in India. However, the “awareness campaign” centric model of cybersecurity strategy does not really work against APT actors, as became apparent soon when researchers at the University of Toronto discovered the most massive infiltration of India’s civilian and military computers by APT actors [4]. In 2013, the Snowden revelations about unrestrained US spying on India also ruffled domestic feathers for lack of any defensive measures or policies [3]. Coupled with these surprise(?) and unpalatable revelations, there was also the increasing and recurring international pressure on Indian IT to put an end to the rising cases of data theft where sensitive data of their overseas customers was regularly found in online underground markets [16].  

    Therefore, with the government facing growing domestic and international pressure to revamp its approach towards cybersecurity, MeitY released India’s first National Cybersecurity Policy in 2013 [17]. Ministry of Home Affairs (MHA) also released detailed guidelines “in the wake of persistent threats” [18]. However, the government admitted to not having the required expertise in the matter, and thus the preparation of the MHA document was outsourced to DSCI. Notwithstanding that, MHA’s document was largely an extension of the Manual on Departmental Security Instructions released in 1994 which had addressed the security of paper-based information. Consequently, the MHA document produced less of a national policy and more of a set of instructions to departments about sanitising their computer networks and resources, including a section on instructions to personnel over social media usage. 

    The 2013 National Cybersecurity Policy proposed certain goals and “5-year objectives” toward building national resilience in cyberspace. At the end of a long list of aims, the 2013 policy suggested adopting a “prioritised approach” for implementation which will be operationalised in the future by a detailed guide and plan of action at national, sectoral, state, ministry, department, and enterprise levels. However, as of this writing the promised implementation details, or any teeth, are still missing from the National Cybersecurity Policy. As continued APT activities [19] show, the measures towards creating situation awareness have also not permeated beyond the technical/collection layer.

    In 2014, the National Cyber Coordination Centre (NCCC) was established, with the primary aim of building situational awareness of cyber threats in India. Given the underwhelming response to the 2013 policy [1], [2], the National Cybersecurity Policy was surmised to be updated in 2020, but as of this writing, the update is still being formulated by the National Cybersecurity Coordinator who heads the NCCC. The present policy gap makes it an opportune subject to discuss certain fundamental issues with cyber situation awareness and the future of cyber defences in the context of the trends in APT activities. 

    Much to Catch Up

    Recently, the Government of India’s Kavach (an employee authentication app for anyone using a ‘gov.in’ or ‘nic.in’ emails-id) was besieged by APT36 [20]. APT36 is a Pak-affiliated actor and what one might call a tier-3 APT i.e., what they lack in technical sophistication, they try to make up for that with passion and perseverance. What makes it interesting is that the malicious activity went on for over a year, before a third-party threat observer flagged it. Post-pandemic, APT activities have not just increased but also shown an inclination towards integrating online disinformation into the malware capabilities [21]. APT actors (and bots), who have increasingly gotten better at hiding in plain sight over social networks, have now a variety of AI techniques to integrate into their command and control – we’ve seen the use of GANs to mimic traffic of popular social media sites for hiding command and control traffic [22], an IoT botnet that had a machine-learning component which the attacker could switch on/off depending upon people’s responses in online social networks [21], as well as malware that can “autonomously” locate its command and control node over public communication platforms without having any hard-coded information about the attacker [23]. 

    Post-pandemic, APT activities have not just increased but also shown an inclination towards integrating online disinformation into the malware capabilities.

    This is an offence-persistent environment. In this “space”, there always exists an information asymmetry where the defender generally knows less about the attacker than the opposite being true. Wargaming results have shown that unlike conventional conflicts, where an attack induces the fear of death and destruction, a cyber-attack generally induces anxiety [24], and consequently, people dealing with cyber attacks act to offset those anxieties and not their primal fears. Thus, in response to cyber-attacks, their policies reflect risk aversion, not courage, physical or moral. It need not be the case if policymakers recognise this and integrate it into their decision-making heuristics. Unfortunately, the National Cybersecurity Policy released in 2013 stands out to be a fairly risk-averse and a placeholder document. Among many other, key issues are: 

    • The policy makes zero references to automation and AI capabilities. This would have been understandable in other domains, like poultry perhaps, but is not even comprehensible in present-day cybersecurity.   
    • The policy makes zero references to hardware attacks. Consequently, developing any capability for assessing insecurity at hardware/firmware levels, which is a difficult job, is also overlooked at the national level itself. 
    • There are several organisations within the state, civilian and military, that have stakes and roles of varying degrees in a robust National Cybersecurity Policy. However, the policy makes zero attempts at recognising and addressing these specific roles and responsibilities, or any areas of overlap therein.
    • The policy does not approach cyber activity as an overarching operational construct that permeates all domains, but rather as activity in a specific domain called “cyberspace”. Consequently, it lacks the doctrinal thinking that would integrate cyber capabilities with the use of force. A good example of this is outer space, where cyber capabilities are emerging as a potent destabiliser [25] and cybersecurity constitutes the operational foundation of space security, again completely missing from the National Cybersecurity Policy.   
    • The policy is also light on subjects critical to cybersecurity implementation, such as the approach towards internet governance, platform regulation, national encryption regime, and the governance of underlying technologies. 

    A Note on the Human Dimension of Cybersecurity

    There exist two very broad types of malicious behaviour online, one that is rapid and superficial, and another that are deep and persistent. The present approaches to building situation awareness in cyberspace are geared towards the former, leading to spatiotemporally “localised and prioritised” assessments [26], matters pertaining to the immediate law and order situations and not stealthy year-long campaigns. Thus, while situation awareness itself is a psychological construct dealing with decision-making, in cybersecurity operations it overwhelmingly has turned into software-based visualisation of the incoming situational data. This is a growing gap that must also be addressed by the National Cybersecurity Policy. 

    The use of computational tools and techniques to automate and optimise the social interactions of a software agent presents itself as a significant force multiplier for cyber threat actors.

    In technology-mediated environments, people have to share the actual situation awareness with the technology artifacts [27]. Complete dependence on technology for cyber situation awareness has proven to be problematic, for example in the case of Stuxnet, where the operators at the targeted plant saw on their computer screens that the centrifuges were running normally, and simply believed that to be true. The 2016 US election interference only became clearer at the institutional level after several months of active social messaging and doxing operations had already been underway [28], and the story of Telebots’ attack on Ukrainian electricity grids is even more telling – a powerplant employee whose computer was being remotely manipulated, sat making a video of this activity, asking his colleague if it could be their own organisation’s IT staff “doing their thing” [29].

    This lack of emphasis on human factors has been a key gap in cybersecurity, which APTs never fail to exploit. Further, such actors rely upon considerable social engineering in initial access phases, a process which is also getting automated faster than policymakers can play catchup to [30]. The use of computational tools and techniques to automate and optimise the social interactions of a software agent presents itself as a significant force multiplier for cyber threat actors. Therefore, it is also paramount to develop precise policy guidelines that implement the specific institutional structures, processes, and technological affordances required to mitigate the risks of malicious social automation on the unsuspecting population, as well as on government institutions.  

    Concluding Remarks

    There is a running joke that India’s strategic planning is overseen by accountants and reading through the document of National Cybersecurity Policy 2013, that does not seem surprising. We have had a troubling policy history when it comes to electronics and communications and are still in the process of shedding our colonial burden. A poorly framed National Cybersecurity Policy will only take us away from self-reliance in cyberspace and towards an alliance with principal offenders themselves. Notwithstanding, an information-abundant organisation like NCCC has undoubtedly to make some choices about where and what to concentrate its attentional resources upon, however, the present National Cybersecurity Policy appears neither to be a component of any broader national security strategy nor effective or comprehensive enough for practical implementation in responding to the emerging threat environment. 

    References

    [1] N. Alawadhi, “Cyber security policy must be practical: Experts,” The Economic Times, Oct. 22, 2014. Accessed: Sep. 14, 2022. [Online]. Available: https://economictimes.indiatimes.com/tech/internet/cyber-security-policy-must-be-practical-experts/articleshow/44904596.cms

    [2] A. Saksena, “India Scrambles on Cyber Security,” The Diplomat, Jun. 18, 2014. https://thediplomat.com/2014/06/india-scrambles-on-cyber-security/ (accessed Sep. 18, 2022).

    [3] C. R. Mohan, “Snowden Effect,” Carnegie India, 2013. https://carnegieindia.org/2013/06/19/snowden-effect-pub-52148 (accessed Sep. 18, 2022).

    [4] R. Dharmakumar and S. Prasad, “Hackers’ Haven,” Forbes India, Sep. 19, 2011. https://www.forbesindia.com/printcontent/28462 (accessed Sep. 18, 2022).

    [5] D. Karthik and R. S. Upadhyayula, “NASSCOM: Is it time to retrospect and reinvent,” Indian Inst. Manag. Ahmedabad, 2014.

    [6] H. C. Fanshawe, Delhi past and present. J. Murray, 1902.

    [7] C. Simms, “Is Social Engineering the Easy Way in?,” Itnow, vol. 58, no. 2, pp. 24–25, 2016.

    [8] J. Lienhard, “No. 1380: Indian telegraph,” Engines Our Ingen., 1998.

    [9] A. Vatsa, “When telegraph saved the empire – Indian Express,” Nov. 18, 2012. http://archive.indianexpress.com/news/when-telegraph-saved-the-empire/1032618/0 (accessed Sep. 17, 2022).

    [10] L. Hoskins, BRITISH ROUTES TO INDIA. ROUTLEDGE, 2020.

    [11] D. R. Headrick, The invisible weapon: Telecommunications and international politics, 1851-1945. Oxford University Press on Demand, 1991.

    [12] B. Parthasarathy, “Globalizing information technology: The domestic policy context for India’s software production and exports,” Iterations Interdiscip. J. Softw. Hist., vol. 3, pp. 1–38, 2004.

    [13] I. J. Ahluwalia, “Industrial Growth in India: Stagnation Since the Mid-Sixties,” J. Asian Stud., vol. 48, pp. 413–414, 1989.

    [14] R. Subramanian, “Historical Consciousness of Cyber Security in India,” IEEE Ann. Hist. Comput., vol. 42, no. 4, pp. 71–93, 2020.

    [15] C. Wiener, “Penetrate, Exploit, Disrupt, Destroy: The Rise of Computer Network Operations as a Major Military Innovation,” PhD Thesis, 2016.

    [16] N. Kshetri, “Cybersecurity in India: Regulations, governance, institutional capacity and market mechanisms,” Asian Res. Policy, vol. 8, no. 1, pp. 64–76, 2017.

    [17] MeitY, “National Cybersecurity Policy.” Government of India, 2013.

    [18] MHA, “NATIONAL INFORMATION SECURITY POLICY AND GUIDELINES.” Government of India, 2014.

    [19] S. Patil, “Cyber Attacks, Pakistan emerges as China’s proxy against India,” Obs. Res. Found., 2022.

    [20] A. Malhotra, V. Svajcer, and J. Thattil, “Operation ‘Armor Piercer:’ Targeted attacks in the Indian subcontinent using commercial RATs,” Sep. 23, 2021. http://blog.talosintelligence.com/2021/09/operation-armor-piercer.html (accessed Sep. 02, 2022).

    [21] NISOS, “Fronton: A Botnet for Creation, Command, and Control of Coordinated Inauthentic Behavior.” May 2022.

    [22] M. Rigaki, “Arming Malware with GANs,” presented at the Stratosphere IPS, Apr. 2018. Accessed: Oct. 19, 2021. [Online]. Available: https://www.stratosphereips.org/publications/2018/5/5/arming-malware-with-gans

    [23] Z. Wang et al., “DeepC2: AI-Powered Covert Command and Control on OSNs,” in Information and Communications Security, vol. 13407, C. Alcaraz, L. Chen, S. Li, and P. Samarati, Eds. Cham: Springer International Publishing, 2022, pp. 394–414. doi: 10.1007/978-3-031-15777-6_22.

    [24] J. Schneider, “Cyber and crisis escalation: insights from wargaming,” 2017.

    [25] J. Pavur, “Securing new space: on satellite cyber-security,” PhD Thesis, University of Oxford, 2021.

    [26] U. Franke and J. Brynielsson, “Cyber situational awareness – A systematic review of the literature,” Comput. Secur., vol. 46, pp. 18–31, Oct. 2014, doi: 10.1016/j.cose.2014.06.008.

    [27] N. A. Stanton, P. M. Salmon, G. H. Walker, E. Salas, and P. A. Hancock, “State-of-science: situation awareness in individuals, teams and systems,” Ergonomics, vol. 60, no. 4, pp. 449–466, Apr. 2017, doi: 10.1080/00140139.2017.1278796.

    [28] “Open Hearing On The Intelligence Community’s Assessment on Russian Activities and Intentions in the 2016 U.S. Elections.” Jan. 10, 2017. Accessed: Dec. 22, 2021. [Online]. Available: https://www.intelligence.senate.gov/hearings/open-hearing-intelligence-communitys-assessment-russian-activities-and-intentions-2016-us#

    [29] R. Lipovsky, “Tactics, Techniques, and Procedures of the World’s Most Dangerous Attackers,” presented at the Microsoft BlueHat 2020, 2020. [Online]. Available: https://www.youtube.com/watch?v=9LAFV6XDctY

    [30] D. Ariu, E. Frumento, and G. Fumera, “Social engineering 2.0: A foundational work,” in Proceedings of the Computing Frontiers Conference, 2017, pp. 319–325.

    [powerkit_button size=”lg” style=”info” block=”true” url=”https://admin.thepeninsula.org.in/wp-content/uploads/2022/10/Research-Paper-TPF-1.pdf” target=”_blank” nofollow=”false”]
    Download
    [/powerkit_button]

  • On Metaverse & Geospatial Digital Twinning: Techno-Strategic Opportunities for India

    On Metaverse & Geospatial Digital Twinning: Techno-Strategic Opportunities for India

    [powerkit_button size=”lg” style=”info” block=”true” url=”https://admin.thepeninsula.org.in/wp-content/uploads/2022/07/TPF_Working-Paper_MetaGDT-1.pdf” target=”_blank” nofollow=”false”]
    Download
    [/powerkit_button]

    Abstract:

    With the advent of satellite imagery and smartphone sensors, cartographic expertise has reached everyone’s pocket and we’re witnessing a software-isation of maps that will underlie a symbiotic relationship between our physical spaces and virtual environments. This extended reality comes with enormous economic, military, and technological potential. While there exist a range of technical, social and ethical issues still to be worked out – time and tide wait for no one is a metaphor well applied to the Metaverse and its development. This article briefly introduces the technological landscape, and then moves over to a discussion of Geospatial Digital Twinning and its techno-strategic utility and implications. We suggest that India should, continue on the existing dichotomy of Open Series and Defence Series Maps, initiate Geospatial Digital Twins of specific areas of interest as a pilot for the development, testing, and integration of national metaverse standards and rules. Further, a working group in collaboration with a body like NASSCOM needs to be formed to develop the architecture and norms that facilitate Indian economic and strategic interests through the Metaverse and other extended reality solutions.

    Introduction

    Cartographers argue that maps are value-laden images, which do not just represent a geographical reality but also become an essential tool for political discourse and military planning. Not surprisingly then, early scholars had termed cartography as a science of the princes. In fact, the history of maps is deeply intertwined with the emergence of the Westphalian nation-state itself, with the states being the primary sponsors of any cartographic activity in and around their territories[1].
    Earlier the outcome of such activities even constituted secret knowledge, for example, it was the British Military Intelligence HQ in Shimla which ran and coordinated many of the cartographic activities for the British in the subcontinent[2]. Thus, given our post-independence love for Victorian institutions, until 2021 even Google Maps had remained an illegal service in India[3].

    One of the key stressors which brought this long-awaited change in policy was the increased availability of relatively low-cost but high-resolution satellite imagery in open online markets. But this remote sensing is only one of the developments impacting modern mapmaking. A host of varied but converging technologies particularly Artificial Intelligence, advanced sensors, Virtual and Augmented Reality, and the increasing bandwidth for data transmission – are enabling a new kind of map. This new kind of map will not just be a model of reality, but rather a live and immersive simulation of reality. We can call it a Geospatial Digital Twin (GDT) – and it will be a 4D artefact, i.e. given its predictive component and temporal data assimilation, a user could also explore the hologram/VR through time and evaluate possible what-if scenarios.

    [powerkit_button size=”lg” style=”info” block=”true” url=”https://admin.thepeninsula.org.in/wp-content/uploads/2022/07/TPF_Working-Paper_MetaGDT-1.pdf” target=”_blank” nofollow=”false”]
    Read the Full Paper
    [/powerkit_button]

  • The Geopolitical Consolidation of Artificial Intelligence

    The Geopolitical Consolidation of Artificial Intelligence

    Key Points

    • IT hardware and Semiconductor manufacturing has become strategically important and critical geopolitical tools of dominant powers. Ukraine war related sanctions and Wassenaar Arrangement regulations invoked to ban Russia from importing or acquiring electronic components over 25 Mhz.
    • Semi conductors present a key choke point to constrain or catalyse the development of AI-specific computing machinery.
    • Taiwan, USA, South Korea, and Netherlands dominate the global semiconductor manufacturing and supply chain. Taiwan dominates the global market and had 60% of the global share in 2021. Taiwan’s one single company – TSMC (Taiwan Semiconductor Manufacturing Co), the world’s largest foundry, alone accounted for 54% of total global revenue.
    • China controls two-thirds of all silicon production in the world.
    • Monopolisation of semiconductor supply by a singular geopolitical bloc poses critical challenges for the future of Artificial Intelligence (AI), exacerbating the strategic and innovation bottlenecks for developing countries like India.
    • Developing a competitive advantage over existing leaders would require not just technical breakthroughs but also some radical policy choices and long-term persistence.
    • India should double down over research programs on non-silicon based computing with a national urgency instead of pursuing a catch-up strategy.

    Russia was recently restricted, under category 3 to category 9 of the Wassenaar Arrangement, from purchasing any electronic components over 25MHz from Taiwanese companies. That covers pretty much all modern electronics. Yet, the tangibles of these sanctions must not deceive us into overlooking the wider impact that hardware access and its control have on AI policies and software-based workflows the world over. As Artificial Intelligence technologies reach a more advanced stage, the capacity to fabricate high-performance computing resources i.e. semiconductor production becomes key strategic leverage in international affairs.

    Semiconductors present a key chokepoint to constrain or catalyse the development of AI-specific computing machinery. In fact, most of the supply of semiconductors relies on a single country – Taiwan. The Taiwan Semiconductor Manufacturing Corporation (TSMC) manufactures Google’s Tensor Processing Unit (TPU), Cerebras’s Wafer Scale Engine (WSE), as well as Nvidia’s A100 processor. The following table provides a more detailed1 assessment:

    Hardware Type

    AI Accelerator/Product Name

    Manufacturing Country

    Application-Specific Integrated Circuits (ASICs)

    Huawei Ascend 910

    Taiwan

    Cerebras WSE

    Taiwan

    Google TPUs

    Taiwan

    Intel Habana

    Taiwan

    Tesla FSD

    USA

    Qualcomm Cloud AI 100

    Taiwan

    IBM TrueNorth

    South Korea

    AWS Inferentia

    Taiwan

    AWS Trainium

    Taiwan

    Apple A14 Bionic

    Taiwan

    Graphic Processing Units (GPUs)

    AMD Radeon

    Taiwan

    Nvidia A100

    Taiwan

    Field-Programmable Gate Arrays (FPGAs)

    Intel Agilex

    USA

    Xilinx Virtex

    Taiwan

    Xilinx Alveo

    Taiwan

    AWS EC2 FI

    Taiwan

    As can be seen above, the cake of computing hardware is largely divided in such a way that the largest pie holders also happen to form a singular geopolitical bloc vis-a-vis China. This further shapes the evolution of territorial contests in the South China Sea. This monopolisation of semiconductor supply by a singular geopolitical bloc poses critical challenges for the future of Artificial Intelligence, especially exacerbating the strategic and innovation bottlenecks for developing countries like India. Since the invention of the transistor in 1947, and her independence, India has found herself in an unenviable position where there stands zero commercial semiconductor manufacturing capacity after all these years while her office-bearers continually promise of leading in the fourth industrial revolution.

    Bottlenecking Global AI Research

    There are two aspects of developing these AI accelerators – designing the specifications and their fabrication. AI research firms first design chips which optimise hardware performance to execute specific machine learning calculations. Then, semiconductor firms, operating in a range of specialities and specific aspects of fabrication, make those chips and increase the performance of computing hardware by adding more and more transistors to pieces of silicon. This combination of specific design choices and advanced hardware fabrication capability forms the bedrock that will decide the future of AI, not the amount of data a population is generating and localising.

    However, owing to the very high fixed costs of semiconductor manufacturing, AI research has to be focused on data and algorithms. Therefore, innovations in AI’s algorithmic efficiency and model scaling have to compensate for a lack of equivalent situations in the AI’s hardware. The aggressive consolidation and costs of hardware fabrication mean that firms in AI research are forced to outsource their hardware fabrication requirements. In fact, as per DARPA2, because of the high costs of getting their designs fabricated, AI hardware startups do not even receive much private capital and merely 3% of all venture funding between 2017-21 in AI/ML has gone to startups working on AI hardware.

    But TSMC’s resources are limited and not everyone can afford them. To get TSMC’s services, companies globally have to compete with the likes of Google and Nvidia, therefore prices go further high because of the demand side competition. Consequently, only the best and the biggest work with TSMC, and the rest have to settle for its competitors. This has allowed this single company to turn into a gatekeeper in AI hardware R&D. And as the recent sanctions over Russia demonstrate, it is now effectively playing the pawn which has turned the wazir in a tense geopolitical endgame.

    Taiwan’s AI policy also reflects this dominance in ICT and semiconductors – aiming to develop “world-leading AI-on-Device solutions that create a niche market and… (make Taiwan) an important partner in the value chain of global intelligent systems”.3 The foundation of strong control over the supply of AI hardware and also being #1 in the Global Open Data Index, not just gives Taiwan negotiating leverage in geopolitical competition, but also allows it to focus on hardware and software collaboration based on seminal AI policy unlike most countries where the AI policy and discourse revolve around managing the adoption and effects of AI, and not around shaping the trajectory of its engineering and conceptual development like the countries with hardware advantage.

    Now to be fair, R&D is a time-consuming, long-term activity which has a high chance of failure. Thus, research focus naturally shifts towards low-hanging fruits, projects that can be achieved in the short-term before the commissioning bureaucrats are rotated. That’s why we cannot have a nationalised AGI research group, as nobody will be interested in a 15-20 year-long enterprise when you have promotions and election cycles to worry about. This applies to all high-end bleeding-edge technology research funding everywhere – so, quantum communications will be prioritised over quantum computing, building larger and larger datasets over more intelligent algorithms, and silicon-based electronics over researching newer computing substrates and storage – because those things are more friendly to short-term outcome pressures and bureaucracies aren’t exactly known to be a risk-taking institution.

    Options for India

    While China controls 2/3 of all the silicon production in the world and wants to control the whole of Taiwan too (and TSMC along with its 54% share in logic foundries), the wider semiconductor supply chain is a little spreadout too for any one actor’s comfort. The leaders mostly control a specialised niche of the supply chain, for example, the US maintains a total monopoly on Electronic Design Automation (EDA) software solutions, the Netherlands has monopolised Extreme UltraViolet and Argon Flouride scanners, and Japan has been dishing out 300 mm wafers used to manufacture more than 99 percent of the chips today.4 The end-to-end delivery of one chip could have it crossing international borders over 70 times.5 Since this is a matured ecosystem, developing a competitive advantage over existing leaders would require not just proprietary technical breakthroughs but also some radical policy choices and long term persistence.

    It is also needless to say that the leaders are also able to attract and retain the highest quality talent from across the world. On the other hand, we have a situation where regional politicians continue cribbing about incoming talent even from other Indian states. This is therefore the first task for India, to become a technology powerhouse, she has to, at a bare minimum, be able to retain all her top talent and attract more. Perhaps, for companies in certain sectors or of certain size, India must make it mandatory to spend at least X per cent of revenue on R&D and offer incentives to increase this share – it’ll revamp things from recruitment and retention to business processes and industry-academia collaboration – and in the long-run prove to be a lot more socioeconomically useful instrument than the CSR regulation.

    It should also not escape anyone that the human civilisation, with all its genius and promises of man-machine symbiosis, has managed to put all its eggs in a single basket that is also under the constant threat of Chinese invasion. It is thus in the interest of the entire computing industry to build geographical resiliency, diversity and redundancy in the present-day semiconductor manufacturing capacity. We don’t yet have the navy we need, but perhaps in a diplomatic-naval recognition of Taiwan’s independence from China, the Quad could manage to persuade arrangements for an uninterrupted semiconductor supply in case of an invasion.

    Since R&D in AI hardware is essential for future breakthroughs in machine intelligence – but its production happens to be extremely concentrated, mostly by just one small island country, it behoves countries like India to look for ways to undercut the existing paradigm of developing computing hardware (i.e. pivot R&D towards DNA Computing etc) instead of only trying to pursue a catch-up strategy. The current developments are unlikely to solve India’s blues in integrated circuits anytime soon. India could parallelly, and I’d emphatically recommend that she should, take a step back from all the madness and double down on research programs on non-silicon-based computing with a national urgency. A hybrid approach toward computing machinery could also resolve some of the bottlenecks that AI research is facing due to dependencies and limitations of present-day hardware.

    As our neighbouring adversary Mr Xi says, core technologies cannot be acquired by asking, buying, or begging. In the same spirit, even if it might ruffle some feathers, a very discerning reexamination of the present intellectual property regime could also be very useful for the development of such foundational technologies and related infrastructure in India as well as for carving out an Indian niche for future technology leadership.

    References:

    1. The Other AI Hardware Problem: What TSMC means for AI Compute. Available at https://semiliterate.substack.com/p/the-other-ai-hardware-problem

    2. Leef, S. (2019). Automatic Implementation of Secure Silicon. In ACM Great Lakes Symposium on VLSI (Vol. 3)

    3. AI Taiwan. Available at https://ai.taiwan.gov.tw/

    4. Khan et al. (2021). The Semiconductor Supply Chain: Assessing National Competitiveness. Center for Security and Emerging Technology.
    5. Alam et al. (2020). Globality and Complexity of the Semiconductor Ecosystem. Accenture.