INTRODUCTION
“So let us embrace technology and AI, but with wisdom and a steadfast commitment to the rule of law,” – Justice Hima Kohli.[1]
The tussle to incorporate established human values through Artificial Intelligence (AI) provided in national constitutions, applied in commercial institutions, and implemented through governmental regulation has always been a concern around the world. It stems from big capitalists to government agencies, who are yet to integrate AI in a controlled manner.[2] Invariably, the concept of the fourth Industrial Revolution advocates that technology-driven measures like AI will play a major role in economic development and trade for aeons to come.[3] Therefore, governments would need to leverage such a revolution, proclaiming their victory even before the start of the race.
FUTURE-PROOFING FTDRA (THE FOREIGN TRADE DEVELOPMENT AND REGULATION ACT, 1992): THE AI ADVANTAGE
The FTDRA (The Foreign Trade Development And Regulation Act, 1992) was enacted to regulate International trade, enabling imports and exports to foster economic development in India.[4] Section 5[5] of the Act mandates that the Act aligns with established Foreign Policies, notified by the Central Government on a periodical basis considering any exceptions, modifications, and adaptations necessary. The Current Foreign Policy 2023 aims at the process of re-engineering and automation to facilitate ease of doing business for exporters based on the principles of trust and partnership. Thus, a fair case of integration of AI is established to honour the pillars on which the Foreign Policy 2023[6] is founded. AI has emerged as a transformative force in trade governance. By acknowledging this change and aligning with the fundamental goals of Foreign Policy, Sections 10[7] and 11[8] serve as the ideal contenders to integrate AI as a proactive strategy to update domestic trade rules and enable smooth trade governance.
Section 10[9] of the Act provides power to the authorized person to inspect goods and services, conduct searches and perform seizures after obtaining due approval of the officer as may be prescribed. However, the process of search and seizure entails many mundane tasks of inspecting every good and service place through human intervention only.[10] This leads to a great amount of time being spent inspecting each shipment, which may rise to large volumes in some cases, leading to human error.
Thus, Seizure Pre-screening through mandatory AI-based Government-certified camera devices, which can pre-screen goods in a search, can be used by authorised persons mandated under Section 10[11] of the Act. This serves as a better alternative to relieve the officer of such menial tasks and restrict him to the larger subset of decision-making. This can also reduce human errors since the manual redundant task of screening the parameters of any goods which may contravene Foreign Policy can be automatically searched and seized expeditiously. An RFID tag can be used to signify that the goods have passed the AI Pre-screening check and do not require further scrutiny. This can help expedite trade and do away with unwarranted searches and seizures.[12] The officer can step in if any grievances arise.
AI is revolutionizing trade, with a prime example being its use in US customs. In US ports, AI conducts pre-screening and searches of goods, aiding officers in expediting inspection processes.[13] Companies are already leveraging AI’s wonders in their supply chains.[14] Therefore, it is imperative that Trade Administrative bodies also derive advantages from this potent technology to maintain efficient welfare governance in trade.[15]
Section 11[16] of the Act talks about the penalties which subsist following any contravention of the provisions mentioned in the Act. Along with the penalties, it also incorporates the process of penalty recovery[17] and settlement, in cases of voluntary admission of contravention by a person in the latter scenario.[18] Such a scenario can benefit from the marvels of AI using an Automatic AI Settlement Process and Penalty Recovery system. An AI-driven chatbot can assist in the process of discussions and formulation of settlement agreements. Upon receiving notification from the Adjudicating Authority, people who acknowledge their contraventions can interact with the chatbot to ascertain, based on specified criteria, the precise settlement amount to be paid.[19]
The AI system can automatically calculate the suitable penalty for exporters by analysing government data on settlements, product value, and acknowledged fines. In cases where the officer must recover the penalty by selling goods for which the penalty has not been paid, automatic auctions through AI can serve the dual purpose of relieving the office of burdensome arrangements of auction, and provide a continuous stream of income for the government. [20]
THE SPECIAL ECONOMIC ZONES ACT, 2005: HARNESSING THE AI ADVANTAGE
Special Economic Zones (SEZs) are specific geographical regions that are set up to attract investors by providing them with benefits such as tax exemptions and duty-free privileges.[21] The proposals for the Act mainly call for the addition of AI in decision-making matters related to SEZs. Section 5[22] of the legislation outlines the guidelines that must be observed when classifying an area as SEZ, with guidelines ranging from the establishment of income-generating prospects to the creation of new employment opportunities. In recent times, the development of new businesses and product lines is finding their origin in private companies through the use of AI technologies, capable of identifying new and profitable business ventures.[23] An exemplary instance of this is the utilization of AI by the multi-billion-dollar corporation, ‘Alibaba’[24], to identify potential profit-generating areas. Applying the same AI assimilation model to Section 5[25] can assist the Government in building new models that can discover new locations for SEZs, exercising guiding parameters.
Additionally, the allocation of loans and grants under Section 35[26] for SEZs using AI as a benefactor in the process can prove vital for the Government in the decision-making process of money allocation. AI can speed up the process of loan/grant analysis and expedite report preparation. This AI integration simplifies allocation administration to SEZs, thereby increasing efficiency and allowing for smoother operations. Additionally, Foreign Trade Policy Specialists could help with the development of AI models wherein they can look at existing models and make new ones that fit National Foreign Trade Policy goals and budget limits. These models would go hand in hand with NITI Aayog’s AI strategy[27], to improve the efficiency of the SEZ structure.
NAVIGATING AI’S TRADE LABRINTH: A MAZE OF PROMISE AND PERIL
Integrating AI into domestic regulation offers a novel approach to addressing global challenges, namely in the realm of trade law, despite its intricate nature. Nevertheless, it is essential to recognize and tackle the related issues, guaranteeing that the Administrative Bodies consider both obstacles and remedies for a thorough integration of AI:
AI’s Black Box Problem: Decision-Making Mystery:[28] The Black Box Problem refers to the inherent problem of defining the relationship between input data and output predictions wherein humans cannot interpret the reasoning behind the output conclusion generated by AI.[29] Every country portrays itself as a culturally sensitive environment where technical studies on the values that underpin various elements of trade should be considered.[30] However, since the rationale behind the output of AI is unknown, it is also difficult to ascertain whether the values in trade are being adhered to by the AI or not. The case of State v. Loomis[31] presents how the use of AI may not necessarily lead to the disclosure of the methodology used in the assessment of a problem despite the algorithm producing satisfactory results for the parties. Thus, meaningful algorithmic auditing[32], by making National AI standards which provide for disclosure of reasoning behind output generation, is the way forward when incorporating AI in trade law.
The Lobbying Conundrum: AI-assisted but Human-centric:[33] The Government’s pragmatic strategy[34] should involve strengthening intergovernmental initiatives[35] to regulate lobbying, particularly among firms that evade compliance.[36] Preventing lobbying excesses can be achieved through the use of human oversight in conjunction with AI-driven administrative decisions.[37] Care should be taken to ensure that lobbying through the use of AI remains in check through logging systems.[38], but also that big firms crucial for AI development, do not leave developing countries due to over-compliance measures.
Addressing Privacy, Algorithms, & IP: Privacy concerns and exploitative use of algorithms[39] and classifications[40] pose significant challenges in trade regulation. India must update trade privacy standards while ensuring compliance with the Digital Personal Data Protection Act, of 2023[41] and GDPR Standards.[42] This would ensure that the Right to Privacy[43] is not violated. Transparency and accountability in decision-making are crucial to addressing algorithmic biases and discrimination in models and safeguarding labour and consumer rights.
CONCLUSION
AI is not just a technological marvel; it is a catalyst for transformation. As India navigates the complexities of integrating AI into trade, it must balance innovation with regulation to unlock AI’s full potential, reshape its trade landscape, and drive economic growth. The future of AI-driven trade regulation is promising, and with strategic implementation, India can lead this transformative journey.
Authors Name: Harjas Singh Gulati & Yash Gupta (NMIMS School of Law, Bengaluru)
References:
[1]Awstika Das, ‘AI is a Game-Changer in Legal Field’: Justice Hima Kohli on Why Artificial Intelligence Does not Pose a Threat, But an Opportunity’ (LiveLaw, 12 February 2023) <https://www.livelaw.in/top-stories/artificial-intelligence-threat-opportunity-game-changer-supreme-court-judge-hima-kohli-221379> accessed 25 February 2024
[2]Matthew U Scherer, ‘Regulating Artificial Intelligence Systems: Risks, Challenges, Competencies, And Strategies’ (2016) 29(2) Harvard Journal of Law & Technology 354 <https://jolt.law.harvard.edu/articles/pdf/v29/29HarvJLTech353.pdf> accessed 27 February 2024
[3]ibid.
[4]The Foreign Trade (Development and Regulation) Act, 1992, No. 22 of 1992, Acts of Parliament, 1992 (India).
[5]The Foreign Trade (Development and Regulation) Act, 1992, §5, No. 22 of 1992, Acts of Parliament, 1992 (India).
[6]ibid.
[7]The Foreign Trade (Development and Regulation) Act, 1992, §10, No. 22 of 1992, Acts of Parliament, 1992 (India).
[8]The Foreign Trade (Development and Regulation) Act, 1992, §11, No. 22 of 1992, Acts of Parliament, 1992 (India).
[9]The Foreign Trade (Development and Regulation Act, 1992 (n 7).
[10]IANS, ‘Large Scale Seizure of Contraband Cigarettes’ (The Times Of India, 9th January 2023) <https://timesofindia.indiatimes.com/business/india-business/large-scale-seizure-of-contraband-cigarettes/articleshow/96854584.cms> accessed 8th March 2024
[11]The Foreign Trade (Development and Regulation Act, 1992 (n 7).
[12]Cogent Infotech, ‘Computer Vision Use Cases in Government Sector’ (Cogent Infotech Blog, 2023) <https://www.cogentinfo.com/resources/computer-vision-use-cases-in-government-sector> accessed 9th March 2024
[13]Department of Homeland Security, ‘Using AI to Secure the Homeland’ (DHS Gov Online, 29th February 2024), <https://www.dhs.gov/ai/using-ai-to-secure-the-homeland> accessed 10th March 2024.
[14]Jennifer Chiamaka, ‘Artificial Intelligence and International Trade Law: Navigating Legal Challenges in the Age of Automation’ (SSRN, 1st January 2023) <http://dx.doi.org/10.2139/ssrn.4706943> accessed 11th March 2024.
[15]ibid.
[16]The Foreign Trade (Development and Regulation) Act, 1992 (n 8).
[17]The Foreign Trade (Development and Regulation) Act, 1992, §11(5), No. 22 of 1992, Acts of Parliament, 1992 (India).
[18]The Foreign Trade (Development and Regulation) Act, 1992, §11(4), No. 22 of 1992, Acts of Parliament, 1992 (India).
[19]Zeleznikow, J., ‘Negotiation, Online Dispute Resolution, and Artificial Intelligence: Handbook Of Group Decision And Negotiation’ (Springer Cham, 11th October 2020) <https://doi.org/10.1007/978-3-030-12051-1_38-1> accessed 12th March 2024
[20]Anonymous, ‘Ai Auction Software’ (AI Auction, 1st March 2024) <https://aiauctionsoftware.com/> accessed 13th March 2024
[21]Buba, Johanne, ‘Special economic zones: an operational review of their impacts’ (World Bank Group, 8th December 2017), <https://documents.worldbank.org/en/publication/documents-reports/documentdetail/3169315 12640011812/special-economic-zones-an-operational-review-of-their-impacts> accessed 14th March 2024
[22]The Special Economic Zones Act, 2005, §5, No. 28 of 2005, Acts of Parliament, 2005 (India).
[23]Michelle Greenwald, ‘How Artificial Intelligence Is Helping Companies Identify & Nail New Product Opportunities’ (Forbes, 17th September 2020), <https://www.forbes.com/sites/michellegreenwald/2020/09/ 17/how-artificial-intelligence-is-helping-companiesidentify–nail-new-product-opportunities/?sh=5f891a5e62d8 > accessed 29th February 2024
[24]ibid.
[25]The Special Economic Zones Act, 2005 (n 22).
[26]The Special Economic Zones Act, 2005, §35, No. 28 of 2005, Acts of Parliament, 2005 (India).
[27]NITI Aayog reports, ‘Responsible AI’ (NITI AAYOG, November 2022), <https://www.niti.gov.in/sites/default /files/2021-02/Responsible-AI-22022021.pdf> accessed 3rd March 2024
[28]Lou Blouin, ‘AI’s Mysterious ‘Black Box’ Problem Explained’ (University Of Michigan-Dearborn 6th March 2023) <https://umdearborn.edu/news/ais-mysterious-black-box-problem-explained> accessed 21st February 2024
[29]Yavar Bathaee, ‘The Artificial Intelligence Black Box And The Failure Of Intent And Causation’ (2018) 31(2) Harvard Journal of Law & Technology 889 <https://jolt.law.harvard.edu/assets/articlePDFs/v31/The-Artificial-Intelligence-Black-Box-and-the-Failure-of-Intent-and-Causation-Yavar-Bathaee.pdf> accessed 28 February 2024.
[30]Markku Suksi, ‘Administrative due process when using automated decision-making in public administration: some notes from a Finnish perspective’ (2021) 29 Artificial Intellignece Law 87, <https://doi.org/10.1007/s10506-020-09269-x> accessed 23rd February 2024
[31]State v. Loomis, 881 N.W.2d 749 (Wis. 2016).
[32]Karen Yeung, ‘Responsibility and AI: Council of Europe Study DGI(2019)05’ (2019) University Of Birmingham <https://rm.coe.int/responsability-and-ai-en/168097d9c5> accessed 24th February 2024
[33]Krish Ramineni, ‘AI Isn’t Here to Take Your Job, but Help You Work Better’ (The Economic Times, 8th February 2024) <https://economictimes.indiatimes.com/jobs/mid-career/ai-isnt-here-to-take-your-job-but-help-you-work-better/articleshow/107526595.cms?from=mdr> accessed 1st March 2024.
[34]NITI Aayog reports, (n 27).
[35]Achar, Sandesh, ‘Early Consequences Regarding the Impact of Artificial Intelligence on International Trade’ (2019) 6(3) American Journal Of Trade And Policy 119 <https://ideas.repec.org/a/ris/ajotap/0133.html> accessed 4th March 2024
[36]I.S. Kim and Helen V. Milner, Global Goliaths: Multinational Corporations In A Changing Global Economy (Vol. 9(4) The Brookings Institution, 2021).
[37]DHC (n 13).
[38]European Commission, Directorate-General for Justice and Consumers, ‘Liability for artificial intelligence and other emerging digital technologies’ (Publications Office, 27th November 2019) <https://data.europa.eu/doi/10.2838/573689> accessed 5th March 2024
[39]Ciuriak, D. and Rodionova, V., ‘Trading Artificial Intelligence: Economic Interests, Societal Choices, and Multilateral Rules’ (Cambridge University Press, 10th May 2021), <https://www.cambridge.org/core/books/artificial-intelligence-and-international-economic-law/trading-artificial -intelligence/D5DC00E20DB7FBA4E4B397CB8AD0C578> accessed 6th March 2024
[40]Vincent J. Straub, Deborah Morgan, Jonathan Bright, Helen Margetts, ‘Artificial intelligence in government: Concepts, standards, and a unified framework’ (2023) 40(4) GOVERNMENT INFORMATION QUARTERLY <https://www.sciencedirect.com/science/article/abs/pii/S0740624X23000813?via%3Dihub> accessed 7th March 2024
[41]The Digital Personal Data Protection Act, 2023, No. 22 of 2023, Acts of Parliament, 2023 (India).
[42]EU GDPR, ‘Eu General Data Protection Regulation Info’ (GDPR, 27th April 2016) <https://gdpr-info.eu> accessed 8th March 2024
[43]Justice K.S. Puttaswamy and Ors. v. Union of India (UOI) and Ors., (2017) 10 SCC 1.
