AI and India’s National Interest
Published By: Hosuk Lee-Makiyama Badri Narayanan Gopalakrishnan Claudia Lozano Dyuti Pandya
Subjects: Digital Economy South Asia & Oceania
Summary
- AI offers a policy dilemma for India. On the one hand, export competitiveness hinges on the nationwide deployment of productivity-enhancing technologies. Indian services and consultancies must incorporate AI technologies to keep their leading position in the global market. On the other hand, there are also concerns over automation’s impact on social cohesion, discriminatory algorithmic decision-making, and the risks of “deepfakes”.
- However, the current transitional guidelines have shown a feasible pathway to avoiding overlapping liability or regulatory blindspots by re-interpreting existing legislation. India has a comprehensive framework for antitrust, corporate liability, free speech, and public order that covers AI development and use cases. India may not need AI-specific rules.
- Currently, only the EU has chosen to legislate through binding laws due to its unique structural deficiencies. The EU lacks a supranational constitution that safeguards human rights and protects citizens against AI-based surveillance or policing by its Member States. The EU must enact binding rules to pre-empt AI laws by national governments that will otherwise fragment its Single Market.
- Whereas India has previously taken inspiration from the EU or the US laws, India must follow its own paths and pursue its national interests based on its services-driven industrial profile and the two-way competition against China and the US, who have chosen not to impose any regulatory impediments on AI use or development.
- India’s challenge lies not in whether businesses prefer a local or foreign AI platform but in encouraging rapid adoption and supporting open-source and other alternatives accessible for fine-tuning and transfer learning for its IT industry.
1. India’s National Interest in AI
Since 2014, Prime Minister Narendra Modi has made it a clear goal to transform India into a digitally connected democracy with the slogan “sabka saath, sabka vikas” or “development for all.” In that policy context, AI is a tool to achieve socio-economic development and pave the way for India to leapfrog on the global stage, aligned with its national economic and political goals.[1]
In parallel, several governmental initiatives (such as the “Aatmanirbhar Bharat” initiative) seek to reindustrialise India and position it as a manufacturing hub,[2] aligning digitalisation with localisation strategies like PMP or “Make it in India” to compete against China and Southeast Asia. A general-purpose productivity tool like AI plays a far more critical role than smartphones, solar panels or e-commerce for industrial policy. Hence, it will be essential for India to strike the right balance on AI to support its service industries.[3]
A recent study assesses India’s AI maturity and readiness as at par with China,[4] and India’s AI market is projected to reach a market size of USD 17 billion by 2027.[5] Competition with China has a significant bearing on India’s industrial policy, as the country is attempting to expand into manufacturing and engineering by leveraging its status as the world’s leading software developer. While India seeks to become an integrated technology supplier, the ICT sector could also surpass the current growth of 7 per cent annually thanks to demographics and geopolitical position.
In other words, India’s export competitiveness hinges on the nationwide deployment of productivity-enhancing technologies. In this rapid adoption, political and social concerns will arise. The European Union (EU) is actively attempting to slow down the global AI race and enacted its EU AI Act after three years of much doubt and intense debate. Elsewhere, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047) passed the State Legislature in California the same year, only to be vetoed by the governor, who cited concerns about high economic consequences from a failed premature legislation.
There are similar sentiments of ambivalence in digital India. The Digital Personal Data Protection Bill (DPDPA) was initially supposed to enter into force by mid-2024, with its publication anticipated by the end of November 2024.[6] While this privacy bill does not explicitly regulate the use of AI, several of its provisions are relevant to AI and challenge the processing of personal data enabled through it. Furthermore, India maintains localisation requirements on data on its telecom users, data on payment and financial transactions,[7] or shareholder information under the Companies Act.[8]
In effect, India already requires foreign developers to de facto conduct the development of language models in India. In addition, there are also new concerns about the impacts automation has on jobs, discriminatory algorithmic decision-making against vulnerable groups, or the risks of “deepfakes” that may threaten social cohesion. Despite the central and state government attempts to fight misinformation and fake news, Indian lawmakers continue to be cautious about the potential risks associated with AI. Notably, during the signing of the Bletchley Declaration in November 2023, then-Minister Rajeev Chandrasekhar stated, “We have learned that by allowing innovation to get ahead of regulation, we open ourselves to toxicity and misinformation.”[9]
The attempts to govern and regulate AI occur as industrialised nations compete globally for industrial leadership. In the same manner that India seeks to re-enter manufacturing, China is building on its engineering prowess to challenge India’s leadership in software. This paper examines how AI entails both opportunities and risks for India. Despite strong macroeconomic fundamentals for growth and favourable geopolitical position positioning, a rapid AI adaptation by commercial rivals like China, the US, and countries in Southeast Asia are forcing Indian IT, finance, and manufacturing industries to be more productive and remain competitive.
[1] Panday and Samdub., Promises and Pitfalls of India’s AI Industrial Policy, AI Nationalisma, 2024, Global Industrial Policy Approaches to AI. Accessed at: https://ainowinstitute.org/wp-content/uploads/2024/03/AI-Nationalisms-Chapter-4.pdf
[2] The Economic Times, Narendra Modi’s “Sabka Saath Sabka Vikas” is great vision: John Kerry, 29 July 2014. Accessed at: https://economictimes.indiatimes.com/news/politics-and-nation/narendra-modis-sabka-saath-sabka-vikas-is-great-vision-john-kerry/articleshow/3921698.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst
[3] Lui, India’s AI Regulation Dilemma, 27 October 2023, The Diplomat. Accessed at: https://thediplomat.com/2023/10/indias-ai-regulation-dilemma/
[4] BSI, Trust in AI, Momentum of AI adoption strongest in India and China, 2024. Accessed at: https://www.google.com/url?q=https://www.bsigroup.com/siteassets/pdf/en/insights-and-media/campaigns/trust-in-ai.pdf&sa=D&source=docs&ust=1730378694159840&usg=AOvVaw0O4oieWj4YcS0uUT7wTZyt
[5] The Economic Times Tech, India’s AI market projected to reach $17 billion by 2027: report, 20 February 2024. Accessed at: https://economictimes.indiatimes.com/tech/technology/indias-ai-market-projected-to-reach-17-billion-by-2027-report/articleshow/107856845.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst
[6] Ministry of Law and Justice, The Digital Personal Data Protection Act, 11 August 2023, MeitY. Accessed at: https://www.meity.gov.in/writereaddata/files/Digital%20Personal%20Data%20Protection%20Act%202023.pdf ; also see: Mookherji, Business Standard, 15 months after DPDP Act, rules to be out by the end of this month, 14 November 2024, Business Standard. Accessed at: https://www.business-standard.com/industry/news/15-months-after-dpdp-act-rules-to-be-out-by-the-end-of-this-month-124111402085_1.html
[7] RBI, Storage of Payment System Data. Accessed at: https://www.rbi.org.in/Scripts/NotificationUser.aspx?Id=11244&Mode=0 ; SEBI, Framework for Adoption of Cloud Services by SEBI Regulated Entities (REs), 2023. Accessed at: https://www.sebi.gov.in/legal/circulars/mar-2023/framework-for-adoption-of-cloud-services-by-sebi-regulated-entities-res-_68740.html
[8] The Insurance Regulatory and Department Authority of India (Maintenance of Insurance Records) Regulations, 2015; Companies Act, 2013 to be read with Companies (Accounts) Rules, 2014
[9] Ministry of Law and Justice, The Digital Personal Data Protection Act, 11 August 2023, MeitY. Accessed at: https://www.meity.gov.in/writereaddata/files/Digital%20Personal%20Data%20Protection%20Act%202023.pdf
2. AI Is a Pillar in India’s Industrial Transformation
The rise of India’s information technology has been a transformative journey and a great success story. While the journey picked pace after India’s economic liberalisation of the 1990s, its origin goes back to before India’s independence. The Indian IT story became one of converting skills and knowledge into wealth and capital.[1] India’s technology development transitioned from state-led economic models to full-scale liberalisation in a relatively brief time. The Indian IT services industry scaled globally, flourishing under private and foreign investment. Thanks to its cost-effective software development and technical expertise, India became renowned as the world’s leading outsourcing partner.
Concurrently, India went from an environment hostile to technology and converted itself into one conducive to change. And this realisation has been crucial in recognising the technology sectors that thrive in India today.[2] Since the Internet revolution in the 1990s, India has added 790 million mobile broadband users, laying the foundation for a massive user base for AI-based services.[3] Analysts expect AI to contribute an additional 1.3% to India’s GDP growth rate by next year, building on the current strong growth. India’s exports are already strongly driven by the domestic technology sector, which accounts for one-quarter of all exports.[4]
The future of India’s IT services is entwined with the nation’s AI development. The integration of AI into Indian IT services and business process outsourcing (BPO) has demonstrated considerable potential to enhance efficiency by automating tasks and client-facing services.[5] Cities like Bangalore risk falling behind their global competitors if they face commercial and regulatory obstacles when using AI. At the same time, foreign competitors invest heavily in AI, making them more cost-effective. Introducing regulatory attempts that can impede AI development in India may allow businesses to repatriate from India and lead to a relocation of IT development and software R&D to other countries with more AI-friendly rules.[6] In other words, Indian IT services and consultancies held back by AI regulations risk losing their hold of the global market.
Fortunately, AI adoption has grown exponentially across industries, with adoption rates around 90% or above among firms in life sciences, financial services in media and telecoms.[7] AI has also brought foreign direct investment (FDI) into free trade zones like the Gujarat International Finance Tec-City. The stakes are high for India as it navigates these dynamics that depend on a competitive business environment to attract FDI.
But India will also need new partners: Computational power for AI development consumes unprecedented amount of electricity, where India’s disadvantaged energy position might become its Achille’s heel. The US and European cloud and infrastructure companies are investing heavily into Saudi Arabia and United Arab Emirates –[8] countries that may be capital and energy rich, but lack India’s human capital, software R&D and experiences in business service.
Given AI’s centrality in the competitiveness narrative, India is among the vast majority of the countries that have taken the “wait and see” approach and avoided introducing binding regulations on its AI developers or use cases of AI, thus taking a different path than the EU and California that have attempted legislation on AI.
[1] Sharma, The Outsourcer: The Story of India’s IT Revolution, 2015, The MIT Press. Accessed at: https://direct.mit.edu/books/book/2242/The-OutsourcerThe-Story-of-India-s-IT-Revolution
[2] Ibid.
[3]Nikekani et al, Unlocking India’s Potential with AI, December 2023, International Monetary Fund. Accessed at: https://www.imf.org/en/Publications/fandd/issues/2023/12/POV-unlocking-india-potential-with-AI-Nilekani-Bhojwani
[4] Nasscom, Technology Sector in India : Strategic Review, 2024. Accessed at: https://nasscom.in/knowledge-center/publications/technology-sector-india-strategic-review-2024
[5] Panigrahi et al, Impact of Artificial Intelligence on Indian economy, Journal of Management Research and Analysis (JMRA), July-September 2024, Accessed at: https://www.jmra.in/html-article/21225#:~:text=In%20addition%20to%20revolutionising%20businesses,growth%2C%20and%20higher%20service%20quality
Prasad Babu, P. and Vasumathi, A,. Role of Artificial Intelligence in Project Efficiency Mediating with Perceived Organizational Support in the Indian IT Sector, Indian Journal of Information Sources and Services, 2023, Indian Journal of Information Sources and Services. Accessed at: https://pdfs.semanticscholar.org/378e/e984d91f50be3e4c970b5da4c2b4295bc8ee.pdf
Nirubarani and Aithal, A Study on the Status of Training in the Indian IT Industry with the Impact of Artificial Intelligence, 2024, Poornaprajna International Journal of Management, Education & Social Science (PIJMESS). Accessed at: https://poornaprajnapublication.com/index.php/pijmess/article/view/22
Shashikala et al,. Role of Artificial Intelligence and its Applications in Indian Banking Sector, May 2024, International Journal of Research Publication and Reviews. Accessed at: https://ijrpr.com/uploads/V5ISSUE5/IJRPR28384.pdf
[6] The Economic Times, Why India risks falling behind in the AI race, 2023. Accessed at: https://economictimes.indiatimes.com/tech/technology/why-india-risks-falling-behind-in-the-ai-race/articleshow/101381214.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst
[7] Ghosh and Mitra, Towards a smarter tomorrow: Impact of AI in the post-COVID era, 2023, PWC. Accessed at: https://www.pwc.in/assets/pdfs/data-and-analytics/towards-a-smarter-tomorrow-impact-of-ai-in-the-post-covid-era-v1.pdf
[8] Alarabiya News, Microsoft to invest $1.5 bln in UAE AI firm, take board seat, 16 April 2024. Accessed at: https://english.alarabiya.net/business/technology/2024/04/16/microsoft-to-invest-1-5-bln-in-uae-ai-firm-take-board-seat-
3. Negative Externalities Cannot Be Fixed with Regulation
While India is not explicitly pursuing ex-ante product regulation on AI akin to those in Europe (or previously planned in California), various agencies have launched conflicting policies, resulting in a minor power struggle that resulted in a fragmented policy landscape. On the one hand, the NITI Aayog is promoting a soft law approach called “Responsible AI” to tackle the ethical, legal, and societal challenges AI technologies pose. In 2018, NITI Aayog began its work on developing guidelines for R&D in agriculture, education and smart cities,[1] followed by a two-part principles document in 2021 for ethical use and self-regulation.[2]
On the other hand, the Ministry of Electronics and Information Technology (MeitY) proposes a comprehensive legislative package to keep pace with AI advancements.[3] In 2022, MeitY announced its plans to replace the Information Technology (IT) Act with the Digital India Act (DIA),[4] where AI applications become defined as intermediaries.
MeitY has also issued a series of advisories, beginning with due diligence requirements for intermediaries regarding content moderation and misinformation under the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules of 2021, a.k.a. Intermediary Rules. The second advisory (issued March 1, 2024) mandated the intermediaries obtain explicit permission to deploy AI models, which attracted public criticism, leading to its revision that withdrew the requirement but instead expanded obligations elsewhere, e.g., under Rule 3 (1) (b) of the IT Act.
On balance, the central government aims to support the monetisation of AI/ML technologies in India while avoiding high-risk use cases and promoting ethical use of AI/ML tools and technology.[5] Some authors claim that there is a threat against the Indian workforce and layoffs caused by automation and AI.[6] However, India has a substantial informal sector and low labour costs that tend to dampen such impacts. Thus, the successful deployment of AI also depends on broader policy initiatives to address systemic inequalities and employment challenges,[7] since simple product regulations cannot fix such structural challenges.
Rather, AI-driven upskilling initiatives in the IT industry are crucial for maintaining a skilled workforce capable of leveraging the technology.[8] Skills deficit and regulatory impediments affect other parts of the Indian economy beyond IT services, including tertiary sectors like telecommunications and banking. In a recent study, 59 per cent of all surveyed enterprises in India had actively deployed AI and started to accelerate their investments in AI and reskill the workforce.
[1] NITI Aayog, National Strategy for AI, 2018. Accessed at: https://www.niti.gov.in/sites/default/files/2023-03/National-Strategy-for-Artificial-Intelligence.pdf
[2] NITI Aayog, Responsible #AIForAll Approach Document for India Part 2 – Operationalizing Principles for Responsible AI, August 2021. Accessed at: https://www.niti.gov.in/sites/default/files/2021-08/Part2-Responsible-AI-12082021.pd . However, the white paper does not define “high risk”
[3] Panday and Samdub., Promises and Pitfalls of India’s AI Industrial Policy, AI Nationalisma, 2024, Global Industrial Policy Approaches to AI. Accessed at: https://ainowinstitute.org/wp-content/uploads/2024/03/AI-Nationalisms-Chapter-4.pdf
[4] Meity, Proposed Digital India Act, 2023 Presentation, 2023. Accessed at: https://www.meity.gov.in/writereaddata/files/DIA_Presentation%2009.03.2023%20Final.pdf
[5] MeitY, Digital India Dialogues held on September 3 2023. Accessed at: https://www.meity.gov.in/writereaddata/files/DIA_Presentation%2009.03.2023%20Final.pdf
[6] Hammer, How Automation Will Shape the Future of Work in India, 27 February 2024, The Diplomat. Accessed at: https://thediplomat.com/2024/02/how-automation-will-shape-the-future-of-work-in-india/
[7] Ibid.
[8] Nirubarani and Aithal, A Study on the Status of Training in the Indian IT Industry with the Impact of Artificial Intelligence, 2024, Poornaprajna International Journal of Management, Education & Social Science (PIJMESS). Accessed at: https://poornaprajnapublication.com/index.php/pijmess/article/view/22
4. Legal System Ready for the Disinformation Challenge
Disinformation (i.e., deliberating creating false information with manipulative intent) and misinformation (when a mob shares disinformation with no intention to deceive) remains one of India’s main concerns with digitalisation. Given India’s online landscape “inundated with manipulated content”,[1] and a massive user base that uses internet services, AI enables psychological profiling and propaganda dissemination with increasing ease to sow confusion and social unrest and disrupt democratic processes. Cases of algorithmic augmentation of misinformation have indeed surged in India: Examples include misleading information during the 2023 Karnataka elections and inflammatory content inciting violence in Manipur,[2] which led the government to temporarily shut down the internet as an emergency measure.[3] Also, earlier this year, Prime Minister Modi raised concerns about speeches by senior leaders of the ruling parties potentially being manipulated by AI.[4]
While the State of California failed to enact a law on AI governance, it has taken legislative action to criminalise deepfakes in political campaigns or sexually explicit ones, including provenance disclosure of AI-generated content to help identify perpetrators. However, California’s state laws were deemed necessary due to limitations of copyright laws (that cannot address cases of abuse in non-commercial contexts) and US free speech protections – and indeed, a federal judge has also temporarily blocked California’s deepfake law citing potential First Amendment violations.
In contrast, India may have specific laws governing deepfakes, but existing laws already offer both civil and criminal reliefs. IT Act of 2000 criminalises identity theft (Section 66C), cheating by personation (66D), and publishing and transmitting obscene material (67 and 67A). Provisions on cyber terrorism (66F) can be invoked against election interference, while authorities may request takedowns of almost any information online.
Provisions under the Bharatiya Nyaya Sanhita, 2023 do not exempt cybercrimes using deepfakes, e.g., sections 189 on speech affecting public tranquillity, 318 on cheating, 319 on cheating by impersonation, 336 on forgery), and 356 on defamation. Further, sections 123 (on major corrupt practices) and 125 (Promoting enmity between classes in connection with election) of the Representation of the People Act, 1951, can also be invoked. Section 51 of the Indian Copyright Act of 1957 may be invoked when copyrighted material is used to create deepfakes.
Admittedly, the challenges posed by deepfake technology in India are particularly complex due to the country’s democratic governance and linguistic diversity. Generative AI has highlighted these unique difficulties, especially in identifying and addressing propaganda and hate speech in various regional languages, making monitoring and enforcement a bigger challenge despite the existing legal basis for prosecution. In sum, a specific deepfake law may not exist in India,[5] but the matter is primarily a question of resources for enforcement rather than the need for new legislation.
4.1. Liability for Effective Accountability for Fake News
In addition to the above, the IT Act contains safe harbour provisions that limit intermediary liability unless the platforms do not remove misleading content from their platforms. In fact, one of MeitY’s motivations for the Digital India Act (DIA) is to expand and reclassify AI platforms as intermediaries based on operational models, assigning both liability limitations and obligations to them.[6] Similarly, the non-binding advisories that were issued by MeitY (as pro tempore measures while drafting the DIA) also target large intermediaries that incorporate AI in their products and reinterpret the due diligence obligations under the IT Act.[7]
For its justification, MeitY argues that the intermediaries do not comply with their due diligence obligations under the IT Act, demanding prompt removal of prohibited deepfakes within 10 hours from notifications in cases that cause irreparable loss and harm to the individuals in instances like cyber frauds and scams.[8]
In other words, India is taking a different path than California’s SB942, which requires generative AI models to include provenance disclosures in created content, i.e. mandates transparency and AI detection tools, thereby facilitating enforcement. Instead, DIA and recent guidelines redefine AI as intermediaries, thereby extending liabilities and limitations on generative AI apps that are not necessarily conventional platforms. But unlike online intermediaries that merely host or transmit third-party content as passive conduits, AI systems actively generate, modify, or curate content. Reclassifying AI as intermediaries extends some safe harbour protections, potentially enabling AI entities to avoid accountability for harmful or biased content their users create. Ultimately, liabilities will end up with distributors of the content, i.e. the traditional online platforms.
To summarise, India has rightly addressed regulatory gaps by reviewing and updating existing legislation rather than carving out AI-specific obligations. While “mere conduits” can implement measures (ironically using AI) to respond to notice and takedown requests, their role is reactive, and transgressors can upload new content faster than they can be detected and removed. Placing liability on platforms fails to address the source of the problem—the unidentified adversaries that use tools and systems that generate the deepfakes or share them. As AI detection technology is still immature, platforms will remove legitimate content to avoid liability, which stifles political freedoms or is likely to backfire on the government.
Focusing liability on distributors shields the developers of generative AI apps, helping them escape accountability unless safeguards like provenance tracking or content authenticity are also implemented – at least until more effective detection algorithms become available. Without holding users and generative AI apps accountable, the root cause of deepfake creation remains unaddressed since deepfake creators use less-regulated intermediaries or foreign jurisdictions that provide no legal assistance treaties.
[1] Ministry of Electronics and IT, Government of India committed towards preventing dissemination of misinformation including deep fakes on Internet, 31 July 2024, Press Information Bureau. Accessed at: https://pib.gov.in/Pressreleaseshare.aspx?PRID=2039640
[2] Varma, 2023 Karnataka Elections: When Polls Made Way for Communal Misinformation, 26 May 2023, The Quint.com. Accessed at: https://www.thequint.com/news/webqoof/post-poll-misinformation-2023-karnataka-elections#read-more
[3] Menon and Manipur: Misleading information shared about India tensions, 26 July 2023, BBC News.com. Accessed at: https://www.bbc.com/news/world-asia-india-66255989
[4] Sebastiant, AI and deepfakes blur reality in India elections, 16 May 2024, BBC News. com. Accessed at: https://www.bbc.com/news/world-asia-india-68918330
[5] Singhania, Navigating the Legal Landscape of AI Deepfakes in India, 2024, Lexolog.com. Accessed at: https://www.lexology.com/library/detail.aspx?g=9157b1f8-4879-4393-89c2-db5238084b1a
[6] Chacko et al, Chambers and Partners, Artificial Intelligence 2024, 28 May 2024. Accessed at: https://practiceguides.chambers.com/practice-guides/artificial-intelligence-2024/india#:~:text=There%20are%20no%20specific%20Indian,intermediary%20liability%20frameworks%20in%20India
[7]Ministry of Electronics and IT, Government of India committed towards preventing dissemination of misinformation including deep fakes on Internet, 31 July 2024, Press Information Bureau. Accessed at: https://pib.gov.in/Pressreleaseshare.aspx?PRID=2039640
[8] MeitY, Merged for MeitY, September 2024 . Accessed at: https://www.meity.gov.in/writereaddata/files/Merged%20for%20MeitY%20website3.pdf
5. Addressing Market Dominance
The dominance of foreign tech giants like Meta, Apple, Google, and Amazon has caused sentiments in India that are echoed from the EU. The prevailing attitude among startups in India is that intermediaries often set the rules of engagement, making it challenging for local startups to compete.[1] The Indian app developers have filed a complaint against Google before the Competition Commission of India (CCI), with more recent instances of Indian startups filing cases against US intermediaries in Madras High Court and Delhi High Court.[2]
The Competition Commission of India (CCI) has also invited proposals to launch a study on the impact of AI on markets and is currently assessing its effect on the antitrust concerns arising from data access, as well as algorithmic biases.[3] Recent prosecutions by the US antitrust regulators against alleged anti-competitive practices have also made India less fearful of retributions for doing the same.[4] In addition, the issue of misinformation and intermediary liability intertwines with AI, as digital platforms have become primary conduits for information dissemination. On the one hand, evermore complex moderation duties favour larger and resourceful platforms before the smaller entities. On the other hand, AI-based moderation can help compliance with notice and takedown requirements.
With past litigations and regulators sharpening their sight on large platforms operating in India, these companies will not leave the crosshairs. While antitrust enforcement entails a company controlling significant market share and abusing that position, AI regulation focuses on regulating developers and users to abide by a normative standard for the technology.
Such regulatory and administrative obligations will not necessarily resolve the core problem of monopolistic business practices because regulations on AI governance can only address issues in the AI development and practical use cases, not monopolistic business practices.[5] Regulating AI will interfere with technological adaptation, which will have undesirable consequences on India’s relative competitiveness. However, it will not change the overall market power of large platforms that stem from unrelated production factors, economies of scale, and cross-subsidisation.
With effective data protection and localisation mandates already in place, market concentration is more effectively addressed with carrots than sticks. Rather than overregulating, mobilising public resources towards local AI models that cater to India’s public interest and local users will be more effective remedy. For instance, the Digital Public Infrastructure (DPI) model may be adapted for AI where data sharing protocols among agencies become an alternative to foreign market structures, similar to how the Open Network for Digital Commerce (ONDC) have sought to decentralise the platform-centric models.[6] Moreover, local initiatives like Bharat Gen that is advancing “sovereign” GenAI, is building localised GenAI and LLMs tailored for a market of 1.3 billion people that reflects Indian diversity.
A “DPI for AI” approach sits with the government’s ambitions of AI optimised for the people of India, but it also entails strategic shift in how India approaches data integration, development and deployment. Despite many achievements to date, India still needs to fully establish a local innovation ecosystem that effectively combines skillsets, R&D, and an ability to scale globally or seek synergies into manufacturing.
[1] Paul, The looming debt inferno: How Indian start-ups are in the grip of a debt trap, 21 January 2024, btMAG.com. Accessed at: https://www.businesstoday.in/magazine/deep-dive/story/the-looming-debt-inferno-how-indian-start-ups-are-in-the-grip-of-a-debt-trap-413501-2024-01-17
[2] Thathoo, 2023 In Review: A Year When Digital Goliath Google Met Its Adversaries In India, 10 December 2023, inc42.com. Accessed at: https://inc42.com/features/2023-in-review-a-year-when-digital-goliath-google-met-its-adversaries-in-india/
[3] Global Legal Insights, AI, Machine Learning & Big Data Laws and Regulations 2024 – India, 2024. Accessed at: https://www.globallegalinsights.com/practice-areas/ai-machine-learning-and-big-data-laws-and-regulations/india/#_ednref31
[4] Kathuria, US has gone after Big Tech. That makes it easier for India, 29 March 2024, The Indian Express. Accessed at: https://indianexpress.com/article/opinion/columns/us-has-gone-after-big-tech-that-makes-it-easier-for-india-9238996/
[5] De Cooman, Without Any Prejudice? The Antitrust Implication of the AI Act 2022, 2022. Accessed at: https://orbi.uliege.be/bitstream/2268/291542/1/De_Cooman_Without-Any_Prejudice_CEUR_Proceedings_2022.pdf
[6]Harmon, et al, Decoding India’s AI Governance Strategy and its Implications for the US-India Bilateral Relationship. Indian Public Policy Review, 2024, Open Network for Digital Commerce. Accessed at: https://ondc.org/about-ondc/
6. Why Europe Needs Binding Laws When India Does Not
In the stalemate of the DIA, it indeed has not escaped the regulators in India that the EU has successfully enacted its AI Act, coming into force in 2025. In the wake of its success, the State of California attempted to follow suit in the regulatory race.
In the most prominent follow-up to the EU, California’s Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047) was meant to regulate what is today deemed “large-scale AI models” involved in humanity-ending disasters involving weapons of mass destruction, mass casualties and other significant physical or economic harm. However, the veto illustrates the difficulties of regulating AI without clearly understanding the risks of premature regulation, potentially misleading the public into thinking that AI is already under effective regulatory control.
The lesson from California shows that lawmakers of other jurisdictions must recognise that the EU differs radically from any other system. For the EU, preventing internal regulatory fragmentation is an important rationale for the AI Act – i.e., to preempt national laws by France, Germany and other Member States that would have otherwise rendered supranational powers moot. Fundamental rights may only be exercised under state laws, with the scope and means of redress varying for each EU Member State.
These structural deficiencies necessitate hard law in the EU. Since each EU Member State is sovereign and the EU lacks an enforceable constitution, the EU AI Act must take the shape of hard law to prohibit its Member State governments from practices like social scoring, discrimination, and biometric identification by law enforcement and migration agencies. This aspect of supranationalism is particularly pertinent when Brussels wants to impose guardrails on several member countries that are veering toward authoritarianism.
In addition, the EU Member States and the institutions agreed to the AI Act right before the EU elections. While the cost of inaction by the European parties or individual politicians would have been high during an election year, EU lawmakers can always escape responsibility for flawed regulations thanks to their multinational and consensus-based collective decision-making. This is a luxury that lawmakers in a real democracy like India do not have.
The EU AI Act is also a product liability regulation typical of Civil Law tradition – i.e., filled with detailed ex-ante obligations tailored to create a specific market outcome rather than avoiding market failure. The EU also diverged from the risk-based approach: Rather than case-by-case decisions subject to proportionality and cost-benefit analysis to avoid overregulation, the EU AI Act designates entire sectors as high-risk, even if the specific uses do not present substantial harm. For instance, not all tasks in the healthcare or banking sector is critical but quite mundane.
Policies modelled after such rigid ex-ante regulation are unsuitable for a fast-evolving technology like AI. Ex-ante regulations are a particularly poor fit for economies under the common law system that seeks to catch up against industrialised countries.
7. Conclusion
India has historically adopted or adapted regulatory frameworks from the EU and the US, often adapting their laws to catch up with technological developments or to align with global standards. The DPDP drew inspiration from EU GDPR, whereas the Competition Act borrowed from the US Sherman Act and its jurisprudence.
However, as India’s digital economy grew and positioned itself as a major technology exporter, its national interests are unique as a nascent industrialised economy that is culturally diverse, and the world’s largest democracy. Moreover, its economic policies must be tailored to its distinct, services-driven industrial profile, demography, and competitive interests.
In conclusion, India’s national interest in AI diverge from Europe and the US in the following ways:
- First, the slow development of digital players in Europe shows that public policy efforts to promote local champions by suppressing foreign providers were ineffective. India’s rapid digitalisation also proves that the private sector does not depend on state-led initiatives to pursue AI or other emerging technologies, as businesses naturally gravitate toward relatively cost-effective software investments with sufficient returns. Free of major regulatory impediments, India’s SMEs and local developers will embrace AI. Many are already doing so due to competitive pressure and intellectual curiosity – not due to regulatory incentives. Europe does not invest in new technologies because its stagnant home market has a profitability problem – which India does not.
- Second, there are similar differences between the governance systems. The EU, unlike India, lacks a robust central authority over national security and fundamental rights, which remain the prerogatives of its member states. The EU must impose binding laws to harmonise market rules and human rights across its Single Market. This is a problem India does not have. Similarly, the balance between free speech and intermediary liability entails entirely different challenges than in the US.
- Third, as a developing country, India is subjected to competition from both Silicon Valley and emerging markets. The challenge lies not in whether businesses will prefer a local or foreign AI platform but whether market regulations will inadvertently stifle India’s momentum. India has already localised a significant portion of the AI value chain, and additional AI-related compliance costs may hamper India’s ability to outpace commercial rivals like China and the US, who have decided to leave AI unregulated.
- Fourth, LLMs are often open-source collaborations or building on scaling, iteration and fine-tuning of open models that benefit India’s software industry. In essence, local IT developers take what Silicon Valley and the global developer community produce (often free of charge) and develop indigenous models commercialised for the domestic or international market. The government of India can use its diplomatic influence to ensure that open-source models remain open, accessible, and commercially viable. The government must continue to foster strategic partnerships for energy security, computing resources, and international standardisation.
In conclusion, regulating too early and against unclear objectives – as in the case of California’s SB 1047 – creates quickly outdated market rules designed to address a limited set of hypothetical risks that we may imagine today. Costs of regulatory failure are too high if India’s exporting capabilities are at stake – as China is increasingly encroaching on India’s software and services dominance by leveraging its prowess in hardware and cloud technologies.
Another lesson from the EU and the US is the need to strengthen and future-proof existing laws rather than produce new ones, especially on corporate liability, consumer protection and free speech, where new laws would disconnect AI use from existing case law and enforcement powers. Similarly, privacy, algorithmic bias, and misinformation are enforcement issues of existing laws rather than appropriate subjects for stand-alone legislation.
Bibliography
Aulakh, (2023). India will regulate AI to ensure user protection, Mint.com. https://www.livemint.com/ai/artificial-intelligence/india-will-regulate-ai-to-ensure-user-protection-11686318485631.html
BSI, (2024). Trust in AI, Momentum of AI adoption strongest in India and China. https://www.google.com/url?q=https://www.bsigroup.com/siteassets/pdf/en/insights-and-media/campaigns/trust-in-ai.pdf&sa=D&source=docs&ust=1730378694159840&usg=AOvVaw0O4oieWj4YcS0uUT7wTZyt
Chacko, M et al., (2024). Artificial Intelligence, Chambers and Partners. https://practiceguides.chambers.com/practice-guides/artificial-intelligence-2024/india#:~:text=There%20are%20no%20specific%20Indian,intermediary%20liability%20frameworks%20in%20India
De Cooman, (2022). Without Any Prejudice? The Antitrust Implication of the AI Act 2022. https://orbi.uliege.be/bitstream/2268/291542/1/De_Cooman_Without-Any_Prejudice_CEUR_Proceedings_2022.pdf
Ghosh and Mitra, (2023). Towards a smarter tomorrow: Impact of AI in the post-COVID era, PWC. https://www.pwc.in/assets/pdfs/data-and-analytics/towards-a-smarter-tomorrow-impact-of-ai-in-the-post-covid-era-v1.pdf
Global Legal Insights, (2024). AI, Machine Learning & Big Data Laws and Regulations 2024 – India. https://www.globallegalinsights.com/practice-areas/ai-machine-learning-and-big-data-laws-and-regulations/india/#_ednref31
Gopalaiah et al., (2024). Deloitte technology trends 2024: India perspective, Deloitte.com. https://www2.deloitte.com/in/en/pages/technology/articles/tech-trends-2024.html
Government of India, (2023). MeitY issues advisory to all intermediaries to comply with existing IT rules, Press Information Bureau. https://pib.gov.in/PressReleseDetailm.aspx?PRID=1990542&ref=static.internetfreedom.in®=3&lang=1
Hammer, (2024). How Automation Will Shape the Future of Work in India, The Diplomat. https://thediplomat.com/2024/02/how-automation-will-shape-the-future-of-work-in-india/
Harmon et al, (2024). Decoding India’s AI Governance Strategy and its Implications for the US-India Bilateral Relationship. Indian Public Policy Review, 5(4 (Jul-Aug))Open Network for Digital Commerce. 2024. https://ondc.org/about-ondc/
Jaishankar and Sirkar, (2024). India’s Tech Strategy: An Introductory Overview, ORF America. https://orfamerica.org/newresearch/india-technology-policy
Kathuria, (2024). US has gone after Big Tech. That makes it easier for India, The Indian Express. https://indianexpress.com/article/opinion/columns/us-has-gone-after-big-tech-that-makes-it-easier-for-india-9238996/
Lui, (2023). India’s AI Regulation Dilemma, The Diplomat. https://thediplomat.com/2023/10/indias-ai-regulation-dilemma/
Marg, (2023). Recommendations on Leveraging Artificial Intelligence and Big Data in Telecommunication Sector, 20 July 2023, Telecom Regulatory Authority of India (TRAI). Accessed at: https://www.trai.gov.in/sites/default/files/Recommendation_20072023_0.pdf
MeitY, (2023). Digital India Dialogues held on September 3 2023. https://www.meity.gov.in/writereaddata/files/DIA_Presentation%2009.03.2023%20Final.pdf
Menon and Manipur, (2023). Misleading information shared about India tensions, BBC News.com. https://www.bbc.com/news/world-asia-india-66255989
Ministry of Electronics and Information Technology, (2024). National Strategy for Artificial Intelligence.https://www.niti.gov.in/sites/default/files/2023-03/National-Strategy-for-Artificial-Intelligence.pdf
Ministry of Electronics and Information Technology, (2024) Government of India committed towards preventing dissemination of misinformation including deep fakes on Internet, Press Information Bureau. https://pib.gov.in/Pressreleaseshare.aspx?PRID=2039640
Ministry of Electronics and IT, (2024). Government of India committed towards preventing dissemination of misinformation including deep fakes on Internet, Press Information Bureau. https://pib.gov.in/Pressreleaseshare.aspx?PRID=2039640
Ministry of Electronics and Information Technology, (2021). Due diligence by the Intermediaries under the Information Technology Act, 2000 and the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, Government of India. https://www.meity.gov.in/writereaddata/files/Merged%20for%20MeitY%20website3.pdf
Ministry of Law and Justice, (2023). The Digital Personal Data Protection Act, MeitY. https://www.meity.gov.in/writereaddata/files/Digital%20Personal%20Data%20Protection%20Act%202023.pdf
Nasscom, (2024). Technology Sector in India: Strategic Review. https://nasscom.in/knowledge-center/publications/technology-sector-india-strategic-review-2024
Nikekani et al., (2023), Unlocking India’s Potential with AI, International Monetary Fund. https://www.imf.org/en/Publications/fandd/issues/2023/12/POV-unlocking-india-potential-with-AI-Nilekani-Bhojwani
Nirubarani, J and Aithal, P. S,. (2024). A Study on the Status of Training in the Indian IT Industry with the Impact of Artificial Intelligence, Poornaprajna International Journal of Management, Education & Social Science (PIJMESS). https://poornaprajnapublication.com/index.php/pijmess/article/view/22
NITI Aayog, (2018). National Strategy for AI #AIForALL. https://www.niti.gov.in/sites/default/files/2019-01/NationalStrategy-for-AI-Discussion-Paper.pdf
NITI Aayog, (2021). National Strategy for AI #AIForALL, 2021. https://www.niti.gov.in/sites/default/files/2021-02/Responsible-AI-22022021.pdf and Part 2.
Panday, J and Samdub, M., (2024). Promises and Pitfalls of India’s AI Industrial Policy, AI Nationalism, Global Industrial Policy Approaches to AI. https://ainowinstitute.org/wp-content/uploads/2024/03/AI-Nationalisms-Chapter-4.pdf
Panigrahi, A, et al,. (2024). Impact of Artificial Intelligence on Indian economy, Journal of Management Research and Analysis (JMRA). https://www.jmra.in/html-article/21225#:~:text=In%20addition%20to%20revolutionising%20businesses,growth%2C%20and%20higher%20service%20quality
Panjiar, (2024). Dealing with deepfakes (MeitY’s Version), Internet Freedom Foundation. https://internetfreedom.in/dealing-with-deepfakes/
Paul, (2024). The looming debt inferno: How Indian start-ups are in the grip of a debt trap, btMAG.com. https://www.businesstoday.in/magazine/deep-dive/story/the-looming-debt-inferno-how-indian-start-ups-are-in-the-grip-of-a-debt-trap-413501-2024-01-17
Prasad Babu, P. and Vasumathi, A,. (2023). Role of Artificial Intelligence in Project Efficiency Mediating with Perceived Organizational Support in the Indian IT Sector, Indian Journal of Information Sources and Services, Indian Journal of Information Sources and Services. https://pdfs.semanticscholar.org/378e/e984d91f50be3e4c970b5da4c2b4295bc8ee.pdf
RBI (2018). Storage of Payment System Data. Accessed at: https://www.rbi.org.in/Scripts/NotificationUser.aspx?Id=11244&Mode=0
Sebastian, (2024). AI and deepfakes blur reality in India elections, BBC News. com. https://www.bbc.com/news/world-asia-india-68918330
SEBI. (2023). Framework for Adoption of Cloud Services by SEBI Regulated Entities (REs). Accessed at: https://www.sebi.gov.in/legal/circulars/mar-2023/framework-for-adoption-of-cloud-services-by-sebi-regulated-entities-res-_68740.html
Securities and Exchange Board of India, (2019). Reporting for Artificial Intelligence AI and machine learning applications and systems offered and used by market infrastructure institutions. https://www.sebi.gov.in/legal/circulars/jan-2019/reporting-for-artificial-intelligence-ai-and-machine-learning-ml-applications-and-systems-offered-and-used-by-market-infrastructure-institutions-miis-_41927.html
Sharma, (2015). The Outsourcer: The Story of India’s IT Revolution, The MIT Press. https://direct.mit.edu/books/book/2242/The-OutsourcerThe-Story-of-India-s-IT-Revolution
Shashikala, K et al,. (2024). Role of Artificial Intelligence and its Applications in Indian Banking Sector, International Journal of Research Publication and Reviews. https://ijrpr.com/uploads/V5ISSUE5/IJRPR28384.pdf
Singhania, (2024). Navigating the Legal Landscape of AI Deepfakes in India, Lexolog.com. https://www.lexology.com/library/detail.aspx?g=9157b1f8-4879-4393-89c2-db5238084b1a
Thathoo, (2023). In Review: A Year When Digital Goliath Google Met Its Adversaries In India, inc42.com. https://inc42.com/features/2023-in-review-a-year-when-digital-goliath-google-met-its-adversaries-in-india/
The Economic Times, (2014). Narendra Modi’s “Sabka Saath Sabka Vikas” is a great vision: John Kerry. https://economictimes.indiatimes.com/news/politics-and-nation/narendra-modis-sabka-saath-sabka-vikas-is-great-vision-john-kerry/articleshow/3921698.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst
The Economic Times, (2023). Why India risks falling behind in the AI race. https://economictimes.indiatimes.com/tech/technology/why-india-risks-falling-behind-in-the-ai-race/articleshow/101381214.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst
Think IAS Think Drishti, (2023). India’s Digital Future: The Digital India Act 2023, How the Digital India Act will shape the future of the country’s cyber landscape. https://www.drishtiias.com/daily-updates/daily-news-editorials/india-s-digital-future-the-digital-india-act-2023
Tripathy, (2024). Analysis – MEITY Advisory On AI Tools And Intermediaries, Mondaq. com. https://www.mondaq.com/india/new-technology/1441712/analysis–meity-advisory-on-ai-tools-and-intermediaries#authors
Varma, (2023). 2023 Karnataka Elections: When Polls Made Way for Communal Misinformation. The Quint.com. https://www.thequint.com/news/webqoof/post-poll-misinformation-2023-karnataka-elections#read-more