AI Monopolies V. Sovereign Nations - Big Tech’s Algorithmic Dominance
Author: LawSphereX
Category: Technology & Cyber Law
Published:
Updated:
AI MONOPOLIES V. SOVEREIGN NATIONS: BIG TECH’S
ALGORITHMIC DOMINANCE
Ankit Raj1
ABSTRACT
Artificial Intelligence (AI) presents transformative opportunities alongside profound ethical and legal challenges, particularly in antitrust domains where Big Tech monopolies leverage AI to entrench dominance, manipulate markets, and challenge sovereign governance. This paper examines the ethical implications of AI, focusing on bias, accountability, and transparency, and evaluates legislative efforts to foster responsible AI development and their governance. It critically analyzes implications of algorithmic collusion, antitrust issues, data access barriers for smaller competitors, and AI-driven strategic manipulation, with a special emphasis on Big Tech’s deep influence.
Case studies of Big Techs illustrate how AIdriven content moderation and market strategies can clash with sovereign regulatory frameworks and eradicate market competition, raising questions about state authority versus tech giants’ power. The paper argues that unchecked AI deployment by Big-Tech risks subjugating democratic governance through opaque algorithms and data monopolies, exacerbating market inequities and societal harms. While global legislative efforts, such as India’s Digital Competition Bill and the EU’s Digital Market Act, GDPR and AI Act, aim to curb these risks, enforcement lags behind AI’s rapid evolution. Therefore, the paper advances a regulatory framework, “Doctrine of Proportionally Smart Governance”, to promote innovation while protecting constitutionalism and safeguarding the concept of equal sovereignty.
Keywords: Artificial Intelligence, Sovereignty, Algorithm, AI Regulation, Anthropic, Cloud Jurisdiction, DPDP Act, AI Act, Dark Pattern, Autonomous Weapon Liability From Disruption to Domination.
1 Ankit Raj, a Delhi based Lawyer, and an Alumni of LC2, Faculty of Law, University of Delhi
Generally, Chess is considered to be a very complex game - played on an 8x8 grid and allows for around twenty possibilities for each individual move. But the game ‘Go’ is played on a 19x19 grid and admits over 200 possibilities per move. The total number of chess games exist are around 10123; but at 10700 the number of potential Go games is at the extent where it can defy comprehension. In 2016, the Artificial Intelligence (AI) product AlphaGo - built by DeepMind2 - defeated the Korean Go champion Leed Sedol3. Before playing this game, AlphaGo was trained by the moves of expert Go players from recorded historical games, resulting in the creation of a database of thirty million moves. What AlphaGo did here was process millions and millions of data points very fast and developed a result intermingling the command and the analysis of data. Such AI is defined as ‘Traditional AI’.
It operates with predefined rules. The next stage of evolution brings the system called ‘Generative AI’, and the most advanced stage of evolution, which is being contemplated, is ‘Agentic AI’4. The final one is expected to operate with autonomy and without predefined rules. Those rules are codes, written in a computer programming language by coders. The rise of artificial intelligence, while majorly being housed by world’s largest technology companies, hasn’t just disrupted industries - it has reordered the set-up of human existence itself. Meta5 gives speech, X6 influences opinions, Amazon7 conducts commerce, and Google8 governs knowledge.
And now we are witnessing that ‘as our independence on digital products, services and platforms grows, coders have become, through design, the issuer of permissions; they are the creators of the new commandments – thou shalt not share information; thou shalt watch this ad before your YouTube video - and the issuers of fatwas, banning famous figures from social platforms for crimes real and imaginary and censoring content’9. The amount of control being wielded by the owner companies of such digital products triggers it to be called ‘algorithmic feudalism’ rather than ‘digital capitalism’. There are allegations 2 DeepMind is a British-American artificial intelligence research laboratory and a subsidiary of Alphabet Inc. , https://deepmind. google/about/, Accessed on June 18th 2025
3 Lee Sedol is a retired South Korean professional Go player of 9-dan rank. Wikipedia page
https://en.wikipedia.org/wiki/Lee_Sedol, Accessed on June 18th 2025
4 McKinsey & Company, When can AI make good decisions? The rise of AI corporate citizens (McKinsey &
Company Article June 4th 2025) https://www.mckinsey.com/capabilities/operations/our-insights/when-can-aimake-good-decisions-the-rise-of-ai-corporate-citizens Accessed on June 19th 2025 5 Britannica, Meta Platforms, https://www.britannica.com/money/Meta-Platforms, Accessed on June 18th 2025
6 Britannica, X, https://www.britannica.com/money/Twitter, Accessed on June 18th 2025
7 Britannica, Amazon, https://www.britannica.com/money/Amazoncom, Accessed on June 18th 2025
8 Britannica, Google, https://www.britannica.com/money/Google-Inc, Accessed on June 18th 2025
9 S Saran and A Sharma, GeoTechnoGraphy
Mapping Power and Identity in the Digital Age (Penguin Random House 2025) 81. against these companies of influencing elections, breaching laws, resisting legitimate authorities, and propagating self-declared rules. 1. The Algorithmic Throne: How Big Tech Entrenches Power The World Economic Forum (WEF) reports that there are now more phones (8. 59 billion) than people (7. 59 billion) on this planet. 10 At the end of 2023, approximately 57% of the global population (4. 6 billion people) were using mobile internet on their own device11. That means the Big Tech companies can get in touch with more than 55% of the world population through their code without the media. Expecting the governments to do the same can be challenging. And that's why the entire fight is about whose law decides what passes through the screen of consumers’ smartphones.
The penetrative tracking systems integrated in the smartphones extract data about locations, food choices, shopping choices, routine, dating preferences, political leaning, philosophical likings and dislikes, and even conversations. The entire recommendation model12 of e-commerce platforms works on individual data. It started with the proposition of enhancing customer experience by customization and quick revenue increase for the companies. But we have hit a point where Alexa is always listening to our conversation, recording it and then the same is being analyzed by a computer. 13 To analyze these human inputs, the AI product development companies use modality-specific models to encode before passing them on to Large Language Model (LLMs)14.
The LLMs broadly take care of following functions: (a) perform sentiment analysis, intent recognition and topic modelling (b) semantic search to match meaning rather than keywords (c) generate text output for the input provider. In case of purchasing, these LLMs predict the items on the basis of your digital profile in the system, whereas in case of social media they assist the overall system to comprehend and map 10 World Economic Forum, “Charted: There are more mobile phones than people in the world”, Published on Apr 11, 2023, https://www. weforum. org/stories/2023/04/charted-there-are-more-phones-than-people-in-theworld/, Accessed on July 8th 2025
11 GSMA*,“The State of Mobile Internet Connectivity 2024” Published In October 2024,
https://www.gsma.com/r/wp-content/uploads/2024/10/The-State-of-Mobile-Internet-Connectivity-Report- 2024.pdf?utm_source=website&utm_medium=button&utm_campaign=somic24, Accessed on June 20th 2025 *GSMA is a global organisation unifying the mobile ecosystem to discover, develop and deliver innovation foundational to positive business environments and societal change
12 Amazon Science, “The history of Amazon's recommendation algorithm” Published on November 22 2019,
https://www. amazon. science/the-history-of-amazons-recommendation-algorithm, Accessed on June 20th 2025 13 New York Times, Amazon’s Alexa Never Stops Listening to You. Should You Worry? Published Aug 8th 2019, https://www. nytimes. com/wirecutter/blog/amazons-alexa-never-stops-listening-to-you/, Accessed on June 19th 14 Large language models (LLMs) are a category of foundation models trained on immense amounts of data making them capable of understanding and generating natural language and other types of content to perform a wide range of tasks. posts according to their meaning and nature, and push them on your (user’s) timeline. This decision of who will be recommended what, both in purchase and in the social media feed, is opaquely governed.
This means the company that controls the algorithm will decide what the buyers buy, what the citizens read, and what they watch about their next Head of State. A Wall Street Journal article flags that spammers are creating an ocean of content using artificial intelligence and Google’s algorithms rank some of those robot-generated pages ahead of the information that you searched for. 15 2. Triad of Democracy, Capitalism and Big Tech Power In February 2021, the Government of India notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules 2021. One of the major clauses was about the identification of the ‘first originator’ of a message when asked by law enforcement in serious cases.
WhatsApp filed a case in the High Court of Delhi against the Government of India16 stating that since the Constitution of India recognises Right to Privacy as a fundamental right (declared by Hon’ble Supreme Court in Puttaswamy v Union of India 2017), this provision of identifying the ‘first originator’ will force WhatsApp to break its end-to-end encryption and breach the privacy of the users. WhatsApp went ahead to the extent of indicating that it will exit India in case it is forced to break its encryption system. 17 It is interesting to note that after the notification of the rule, in May 2021, the privacy policy of WhatsApp was updated. The policy controlled the access to the most personal information of the users, and WhatsApp was allowed to publish the same in every fashion.
18 Therefore, it is obvious that WhatsApp was willing to assume the role of custodian of the people of India's right to privacy, but it resists the Government, which is the actual mandate holder, directly from the people of India. This is possible because WhatsApp has gained significant
15 Wall Street Journal, Beware the Top Google Search Result. It Might Be Wrong, February 18th 2024,
https://www.wsj.com/tech/ai/google-search-results-ai-spam-cf8c2605, Accessed on June 21st 2025
16 WhatsApp LLC v Union of India W.P. © 5480/2021; Pending before the High Court of Delhi
17 Live Mint, Will cease to function if forced to break encryption, WhatsApp tells Delhi High Court, April 25th 2024, https://www.livemint.com/companies/news/will-cease-to-function-if-forced-to-break-encryption-whatsapp-tellsdelhi-high-court-11714047728696.html, Accessed on June 20th 2025
18 SCC Online, WhatsApp v Right to Privacy
Supreme Court directs WhatsApp to publicize its May 2021 undertaking, Published on Feb 3rd 2023, https://www. scconline. com/blog/post/2023/02/03/directed-whatsappto-widely-publicise-stand-that-its-users-in-india-do-not-have-to-accept-its-2021-privacy-policy-in-order-to-usemobile-application/, Accessed on June 20th 2025 power by having more than 400 million users in India. 19 This is a case where power has been gained by creating almost a monopoly. The other example is of X (formerly Twitter) where the owner of the company, Elon Musk, funded the Presidential candidate’s election campaign, and when the same candidate, Donald Trump, won and assumed the office of the President, Musk secured government contracts worth billions of dollars under the aegis of his company SpaceX.
20 This stands the case where Big Tech gets power not initially by quantum of users but by influencing elections, funding the potential decision makers in the Government, and reaping private profit from biased decisions of public office holders. Both the phenomena, one where the Government of India failed to make WhatsApp fall in line with its constitutionally valid law, and the other one is Elon Musk’s case where the Government of USA acted in a biased manner and Musk’s rocket company SpaceX got a disproportionate number of federal contracts in USA; are direct threats to democracy. A less politically charged example which underlines the Big Tech power is that of the techgiant company ‘Microsoft’.
On July 19 2024, due to the outage caused by a defect in the update of Microsoft software ‘Windows’, various services like civil aviation, train services, stock exchanges, banks, supermarkets, healthcare systems, and emergency services went defunct not in any specific country rather across the globe including India, UK, USA and other nations. 21 We live in a world where Indians travel 25 million Kilometers on Uber cabs per day making a total of 9. 2 billion kilometer rides in 2024. 22 One day shutdown of Uber can bring cities to halt causing discomfort and economic loss to millions of individuals. And by virtue of its scale, it will become a problem for the government. 19 The Economic Times, WhatsApp Channels surpasses 500 million monthly active users, Published on Nov 15th 2023, https://economictimes. indiatimes.
com/tech/technology/whatsapp-channels-crosses-500-mn-monthlyactive-users/articleshow/105241047. cms? from=mdr, Accessed on June 20th 2025 20 New York Times, Musk Is Positioned to Profit Off Billions in New Government Contracts, Published on March 23 2025, https://www. nytimes. com/2025/03/23/us/politics/spacex-contracts-musk-doge-trump. html, Accessed on June 20th 2025 21 The Economic Times, Microsoft CrowdStrike: How the massive outage is affecting flights and services in India and beyond, July 20th 2024, https://economictimes. indiatimes. com/industry/transportation/airlines-/- aviation/microsoft-crowdstrike-how-the-massive-outage-is-affecting-flights-and-services-in-india-andbeyond/articleshow/111864652. cms? from=mdr, Accessed on June 20th 2025 22 Business Standard, Kochi to Ayodhya: India Ubered its way to 9. 2 billion kilometres in 2024, January 9th 2025, https://www.
business-standard. com/companies/news/uber-2024-highlights-rider-rating-kochi-ayodhyadelhi-mumbai-bengaluru-125010900691_1. html, Accessed on June 21st 2025 These deep entrenchments of such companies underline the new realities of power balance. Despite not liking it, the governments are forced to share the sovereignty floor. This also seems to be a reverse of the theory of New Public Management23 articulated in the 1980s, propagating the thought of “running the government like business” because today the Big Tech companies seem to be somewhat running like the government.
Today Amazon Web Services (AWS) provides safe cloud to the United States Intelligence agency24; despite the fact that India and China are the two most populated nations of the world, the data centers available in both the countries respectively 153 and 449 is way lesser than that of United States’ 542725 resulting in vast amount of data of the individuals of these countries travelling to USA and stored in private companies facilities26. The phenomena of platform monopoly can be understood by the fact that Apple, the largest company on earth by market capitalization of 3. 14 Trillion United States Dollar (USD), charges 30% for hosting paid services of mobile apps on it Appstore, and the same has been prima facie found to be abuse of its dominant position by Competition Commission of India27.
When it comes to digital payments on Unified Payment Interface (UPI) in India, ‘Google Pay’ and ‘Phone Pay’ almost enjoy a duopoly by capturing a whopping 85% market share of digital payments despite not being banks28. 3. The Digital Sovereignty Crisis: Law without Borders, Borders without Law It is noteworthy that medieval trade fairs were not substitutes for the institutions of feudal authority. The merchants then developed Lex Mercatoria (Merchant Law), a private set of rules, which governed their conduct of business. 29 We see a somewhat repetition of the same in today’s Big Tech’s functioning. They don't really erode the conventional sovereignty of the State rather add a layer of their own sovereignty that remains away from the control of the State. 23 Britannica, New Public Management, https://www. britannica.
com/topic/governance/The-new-publicmanagement Accessed on June 20th 2025 24 Amazon, Cloud Computing for U. S. Intelligence Community, https://aws. amazon. com/federal/us-intelligencecommunity/, Accessed on June 20th 2025 25CloudScene, Data Centers in the World, https://cloudscene. com/region/datacenters, Accessed on June 20th
26 BlackBridge Research and Consulting, Top 10 Data Center Companies in USA,
https://www.blackridgeresearch.com/blog/top-data-center-companies-in-usa-united-states, Accessed on June 20th 2025 27 Competition Commission Of India, Case No. 24 of 2021, Together We Fight Society vs Apple Inc. & Another 28iNC 42, PhonePe, Google Pay Lead UPI Race In May, June 12th 2025, https://inc42.com/buzz/phonepegoogle-pay-lead-upi-race-in-may/, Accessed on June 21st 2025
29 J.S. NYE (Jr.), The Future of Power, (PublicAffairs 2021,1st ed), 119.
Edward Snowden30, a cryptocurrency tycoon and whistleblower, explains his idea of struggle about Digital Sovereignty in a speech as “Sovereignty of the people Versus Sovereignty of the State”. 31 The act of Twitter blocking Donald Trump’s Twitter account in 2021 citing its public interest framework as basis to mention “we made it clear going back years that these accounts are not above our rules and cannot use Twitter to incite violence”32. In October 2021 Twitter admitted33 that its algorithm is not neutral, but rather biased. It has been proven by research that Twitter favors the tweets of one side of the political spectrum over the others. 34 After buying Twitter, in 2022, the new CEO Elon Musk retrieved Donald Trump’s account.
Now in hindsight one may ask a question - “Who is governing the principles and policies of a company which has unparalleled size of users and have deep impact on the freedom of expression of 429 million35 people across the world? ” ‘The root of the problem lies in trying to impose the rules of national sovereignty onto a virtual space that inherently defies these rules. The problem is compounded manifold when corporations begin to act like sovereign entities with the whole world as their fiefdom, setting up artificial barriers where none should exist. ’36 This paradigm ignores the understanding that in various parts of the world the constitutional concept and rights are processed, interpreted and enjoyed to different degrees. Free speech and permissible speech are not to the same degree in all national jurisdictions.
This can be understood by the comparison of the interpretation of the First Amendment of the American Constitution and that of Article 19 of the Constitution of India. The First Amendment37 does not recommend any restriction and there is an extremeness in its nature where citizens can say whatever they want, and the Government of the USA cannot 30 Britannica, Edward Snowden, Updated on July 21st 2025, https://www. britannica. com/biography/Edward- Snowden, Accessed on August 8th 2025
31 Edward Snowden, [REDACTED] 2024 | Digital Sovereignty in the Age of AI With Edward Snowden
(YouTube, November 12 2024, at 3:02) https://youtu.be/VCfTLQo5QZ0?si=4uBNerokr9EJaHV0&t=199 accessed 20th June 2025. 32 BBC, Twitter 'permanently suspends' Trump's account, January 9th 2021, https://www.bbc.com/news/worldus-canada-55597840, Accessed on June 20th 2025
33 BBC, Twitter's algorithm favours right-leaning politics, research finds, 22nd October 2021,
https://www.bbc.com/news/technology-59011271, Accessed on June 20th 2025 34 F Huszár, Algorithmic Amplification of Politics on Twitter, 2021 https://cdn.cmstwdigitalassets.com/content/dam/blog-twitter/official/en_us/company/2021/rml/Algorithmic-Amplification-of- Politics-on-Twitter.pdf, June 20th 2025
35 Statista, X (formerly Twitter - Statistics and facts), October 16th 2024,
https://www.statista.com/topics/737/twitter/#topicOverview, Accessed on June 20th 2025 36 S Saran and A Sharma, GeoTechnoGraphy: Mapping Power and Identity in the Digital Age (Penguin Random House 2025) 37 The First Amendment of the Constitution of USA is “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.” https://constitution.congress.gov/constitution/amendment-1/, Accessed on June 21st 2025 put any restrictions on. The degree of this freedom is different in India. Article 1938 in it Clause
1 guarantees various freedoms, but in very immediate Clause (2) prescribed reasonable
restrictions. It has been held that the State cannot travel beyond the contours of clause (2) to (6) of Article 19 of the Constitution of India in curbing the fundamental rights guaranteed by clause (1). The court generally interprets statutes with the principle of Ut Res Magis Valeat Quam Pereat39 to give them meaning and not destroy them, by way of purposive interpretation and constitutional morality, goes the extra mile to secure the fundamental rights from the power of the state. It has been held that in case of reasonable restrictions on the fundamental rights as guaranteed in Article 19(1), the court is not concerned with the necessity of the impugned legislation or the wisdom of the policy underlying it, but only whether the restriction is in excess of the requirement and whether the law has overstepped the constitutional limitation.
40 Taking authority from this grundnorm, when the Ministry of Electronics and Information Technology drafted Intermediary Rules (IT rules) of the Information Technology (IT) Act to govern the operation of social media platforms like Twitter, in 2023, upon Twitter’s resistance to comply the then Hon’ble Union Minister had to say that “Twitter behaved as if national [Indian] laws did not apply to it”41 Under Title III, the French Constitution of 1791, says “Sovereignty is one, indivisible, inalienable, and imprescriptible. It appertains to the nation; no section of the people nor any individual may assume the exercise thereof.
”42 The United Nations (UN) Charter follows the principle of ‘Sovereign Equality’ for all its member states and has enumerated in Article 2(1) of the UN Charter “The Organization [UN] is based on the principle of the sovereign equality of all its Members [Nation States]. ”43
38 Article 19 of the Constitution of India “(1) All citizens shall have the right—
(a) to freedom of speech and expression; (2) Nothing in sub-clause (a) of clause (1) shall affect the operation of any existing law, or prevent the State from making any law, in so far as such law imposes reasonable restrictions on the exercise of the right conferred by the said sub-clause in the interests of the sovereignty and integrity of India, the security of the State, friendly relations with foreign States, public order, decency or morality, or in relation to contempt of court, defamation or incitement to an offence. 39 Ut res magis valeat quam pereat - meaning “that the thing may rather have effect than be destroyed” - reflects the judicial principle that statutes should be interpreted to give them meaning and efficacy rather than rendering them void. See A Fellmeth and M Horwitz, Guide to Latin in International Law (Oxford University Press, 2022,
2 ed.)
40 M P Jain, Indian Constitutional Law, (LexisNexis, Eighth edition, 2018), 1052.
41 Rajeev Chandrashekhara, Post on X (Formerly Twitter) June 13th 2023,
https://x. com/RajeevRC_X/status/1668455627319033857 , Accessed on June 21st 2025 42Original Sources, History of France, https://www. originalsources. com/Document. aspx? DocID=ZVK21Q2EQ91NQDA Accessed on June 21st 2025 43 UN Charter, Chapter - 1 Purposes and Principles, https://legal. un. org/repertory/art2. shtml, Accessed on June The ratification of United Nations policies has brought a certain degree of socio-legal homogeneity. Liberalisation of markets and integration of supply chains seeded the codependency. And the recent phenomena of digitalisation powered by artificial intelligence, machine learning, augmented reality and virtual reality have democratised the homogeneity formulated by strong corporate companies. This prepares the ground for such companies to wield economic, strategic, technological, and cultural power to bend others to their sovereign will. 4.
The Global Regulatory Chessboard Generally a software application (app) either functions as a junction for service consumers and service providers to connect each other as in case of Uber44, Myntra45, Amazon etc. or it functions like a service provider of the company which has developed it, as in case of X, Instagram, WhatsApp etc. For using each app, the user has to mandatorily license the app to access to his/her digital personal data, activities and continuous tracking; in return the app provides license to the user to access the services or products provided by the company through the app. And the legal document that facilitates this process is the End User License Agreement (EULA). Amazon collects various types of data46 about the individual, including the choices of goods, credit ratings, corporate ratings, voice, locations etc.
and provides it to the third party which either integrates its services to Amazon ecosystem or works on behalf of Amazon. In case that third party company is sold to another company, the buyer company gets that data as an important asset of the overall acquisition. Amazon sells its products on its own marketplace platform and has been facing allegations of violation of competition law. 47 Anthropic48 was issued with a government order by the Government of USA to stop access of its most effective models Fable 5 and Mythos 5 of its AI 21st 2025 44 Uber is an American company that provides services related to mobility. The business’s signature product is an application for smartphones that allows users to request transportation from drivers of privately owned vehicles.
45 Myntra is India's largest e-commerce store for fashion and lifestyle, offering a wide range of brands, and authentic products
46 Amazon, Amazon Privacy Notice, 2025
https://www.amazon.in/gp/help/customer/display.html?nodeId=GX7NJQ4ZB8MHFRNJ, Accessed on June 24th 47Hausfeld, E-commerce giant Amazon faces legal action for unlawfully favouring its own product offer, October 20th 2022, https://www.hausfeld.com/news/e-commerce-giant-amazon-faces-legal-action-for-unlawfullyfavouring-its-own-product-offers, Accessed on June 25th 2025
48 https://www.anthropic.com/, Anthropic is a US based AI company, Accessed on June 14th 2026
tool Claude for all non-American nationals. Anthropic issued an official statement that it will eventually stop the access of all customers to those models49. Amidst all these, the question for judicial examination is whether that data becomes public property, and gets exempted from the sphere of right to privacy of the individual, by the time it reaches third or fourth hand? Also can this data be used by the companies without any limit on time and purpose? To what extent can the government trust these companies to self-govern themselves? This is where regulation, judicial interpretation and concept of sovereignty converge.
The European Union’s (EU) General Data Protection Regulation (GDPR), 2018 being the foremost major legislation, inspires jurisdictions across the world with its provisions based on principles like Data Protection by Design and Default and heavy penalties in case of breach. The Digital Services Act (DSA), 2022 of the EU focuses on algorithmic transparency of Very Large Online Platforms (those who have more than 45 million users). The Digital Markets Act (2022) of the EU focuses on regulating the platforms which have strong market power and large user base from self-preferencing and discriminatory gatekeeping. The Data Governance Act (2022) of the EU focuses on safe and secure sharing of data via trusted data intermediaries. The EU’s Data Act (2023) targets to ensure that users have access to the data collected from various devices that they use.
The Online Safety Act (2023) of the United Kingdom (UK) seeks to impose a duty of care on online platforms to protect users from harmful content. The Personal Information Protection Law (PIPL) of China ensures that strong restrictions are enforced on sensitive and cross border data. India’s Digital Personal Data Protection (DPDP) Act (2023) makes provisions for individuals to have access to the data processed, for data fiduciaries to erase data of individuals when its no more required, increased control of Central Government on cross border data sharing and heavy penalties for companies in case of non-compliance. The EU’s AI Act 2024 classifies AI into four risk-categories where predictive policing using sensitive data remains banned by virtue of being classified in the ‘Unacceptable risk’ whereas AI in gaming remains at the lowest risk category.
The Data (Use and Access) Act 2025 in UK provisions for giving more control to individuals as to what data they want to share with the accredited third parties, expands government’s jurisdiction for accessing the data on grounds like crime, national security threat etc, at the same time it also makes it easy for data to travel 49 https://www. anthropic. com/news/fable-mythos-access - Anthropics Statement about compliance with US Government order regarding access to Fable 5 and Mythos 5 models of Claude AI, Accessed on June 14th 2026 cross border without the destination mirroring a UK-GDPR regime. Most of these laws aim to regulate the levers of the data ecosystem after realizing the issues on the strategic, economic, legal or sovereignty fronts. But the speed at which the AI ecosystem and machine learning technology is evolving, these laws are vulnerable to be bypassed.
The core question “Who owns the data of individuals once the company collects it? ” has remained unanswered. In India, the Right to Privacy is read under Article 21 of the Constitution, is a ‘Right to be let alone’. A person has a right to safeguard the privacy of his own. None can publish anything concerning the above matters without his consent50. In the light of this observation by the Hon’ble Supreme Court of India, can this selling or transfer of personal data by the digital platforms to third parties and then subsequent exchange be qualified to be called equal to “publishing without consent” and hence violation of Right to Privacy? Or it can be covered under the “fair use” principle.
The judicial examination of the commerce of data is crucial, keeping in view the recent judgment of the United States District Court Northern District Of California (June 23rd 2025) where Anthropic ABC an AI software firm bought millions of copyrighted books from pirate sites online and feed the entire content in its system to train it. The court's decision establishes that even if an AI system uses the data from a pirate market for training its LLMs, the ownership of data does not change, and the individual, to whom the data belongs, remains the sole owner. 51 5. Reclaiming the Digital Republic Steve Jobs once said “Computers have always been a bicycle of the human mind”. 52 Scholars have already proved that when an LLM gets trained from the data extracted from another LLM, there are high chances of model collapse.
53 When it comes to the presence of competition in the Data market, the world is witnessing intense concentration. We have already seen the example of control of Meta on the social media market. In a contrary theory David Author says “AI is a tool, like a calculator or a chainsaw, and tools generally aren’t substitutes for expertise but rather levers for its application. ” A recent example of the dependency of AI systems on human
50 UPSC v. RK Jain, (July 13, 2012)
51 ANDREA BARTZ, CHARLES GRAEBER, and KIRK WALLACE JOHNSON, v. ANTHROPIC PBC,
UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA
52 YouTube, Steve Jobs explains why computers are a bicycle for the mind
“We humans are tool builders”, April 27th 2024, https://youtu. be/NjIhmzU0Y8Y? si=0alkXOFpSvj7S2V4, Accessed on June 26th 2025 53 Shumailov, I. , Shumaylov, Z. , Zhao, Y. et al. AI models collapse when trained on recursively generated data. Nature 631, 755–759 (2024). https://doi. org/10. 1038/s41586-024-07566-y, Accessed on June 26th 2025 expertise is of a financial technology (fin-tech) company which went very aggressive on AI adoption in 2023 is back to human hiring to run the company. 54 At the same time we are also witnessing Meta unveiling a USD 14. 3 billion investment and partnership that will be the core of a new artificial intelligence research lab dedicated to the pursuit of “superintelligence”. 55 In a rapidly evolving paradigm where the global market for Synthetic Data Generation was estimated at USD 323. 9 Million in 2023 is projected to reach US$3.
7 Billion by 2030, growing at a Compound Annual Growth Rate (CAGR) of 41. 8% from 2023 to 2030. 56 58* Apart from this massive economic opportunity, the world is keenly interested in how the AI Products transform and evolve, i. e. from Traditional AI to Generative AI to Agentic AI59 to
54 Bloomberg, Klarna Slows AI-Driven Job Cuts With Call for Real People, May 8th 2025
https://www.bloomberg.com/news/articles/2025-05-08/klarna-turns-from-ai-to-real-person-customer-service, Accessed on June 26th 2025
55 New York Times, Meta Is Building a Superintelligence Lab. What Is That?, June 13th 2025,
https://www.nytimes.com/2025/06/13/technology/meta-scale-ai-super-intelligence-lab.html, Accessed on June 26th 2025
56 Businesswire, Research and Market Report, January 13th 2025,
https://www.businesswire.com/news/home/20250113130135/en/Synthetic-Data-Generation-Business-Research- Report-2024-Global-Market-to-Reach-%243.7-Billion-by-2030-from-%24323-Million-in-2023-Driven-by- Rising-Demand-for-Data-Privacy-and-Anonymization-Solutions--- ResearchAndMarkets.com?utm_campaign=shareaholic&utm_medium=copy_link&utm_source=bookmark
57 Ibid.
58 EdgeDelta, Breaking Down The Numbers
How Much Data Does The World Create Daily in 2024?, March 11th 2024, https://edgedelta.com/company/blog/how-much-data-is-created-per-day/, Accessed and analysed on June 26th 2026 * 1 Zettabyte = 1 Trillion Gigabytes (GB)
59 McKinsey & Co., When Can AI make good decisions?,June 4th 2025,
https://www. mckinsey. com/capabilities/operations/our-insights/when-can-ai-make-good-decisions-the-rise-of-ai- Physical AI60. The amount of data being created has noticed an exponential increase. 61* This underlines the unconventional nature and speed of evolution and hence requires unconventional measures from the legislators as well. It is hard to expect that the law makers (both who understand the technicality of this technological evolution and those who don’t) will be able to keep up with its governance as they deal with conventional issues. AI developed and currently governed by Big Tech companies impacts not only the commercial ecosystem but also the very core of democracy - fundamental and legal rights, elections and security of institutions and individuals.
The flip side of its governance is the potential of AI to facilitate innovation, competition across industries, and that requires that the Big Techs are appropriately regulated. To that purpose, it is proposed that we adopt the ‘Doctrine of Proportionally Smart Governance’. corporate-citizens, Accessed on June 26th 2025
60 Sierra Ventures, The Future of Work
From GenAI to Agentic AI to Physical AI, January 28th 2025 https://www.sierraventures.com/content/the-future-of-work-from-genai-to-agentic-ai-to-physical-ai, Accessed on June 26th 2025
61 EdgeDelta, Breaking Down The Numbers
How Much Data Does The World Create Daily in 2024? , March 11th 2024, https://edgedelta. com/company/blog/how-much-data-is-created-per-day/, Accessed and analysed on June 26th 2025 * 1 Zettabyte = 1 Trillion Gigabytes (GB) 6. Doctrine of Proportionally Smart Governance The governance architecture for Big Tech and AI must be calibrated to match the pace, complexity, and risk profile of the technology it regulates. Achieving this requires deliberate institutional redesign within the legislative process itself, not incremental amendment of existing frameworks, but the creation of dedicated structures capable of sustained technical engagement. This entails three foundational requirements.
First, a permanent, statutorily-mandated technical cadre embedded within the policy formulation process, with defined terms of reference, reporting lines, and decision-making authority - distinct from ad hoc advisory committees. Second, a shift toward principle-based, technology-neutral drafting that anticipates foreseeable technological trajectories rather than codifying present-state assumptions that obsolesce within a legislative cycle. Third, a structured, time-bound multi-stakeholder consultation process - covering citizens, civil society, domestic industry, and global technology firms - built into the rule-making timeline as a mandatory procedural step, not a discretionary one.
These structures exist to reconcile three objectives that frequently come into tension: the protection of fundamental rights (privacy, autonomy, non-discrimination), the preservation of competitive market conditions for both domestic and foreign players, and the avoidance of compliance burdens disproportionate to firm size or risk posed. On this foundation, the following legislative actions are proposed:
1 Algorithmic Manipulation and Consumer Choice Act
Enact a standalone statute prohibiting manipulative algorithmic design practices in consumer-facing digital products. The statute should codify an enforceable, technically-specified right for users to opt out of personalised targeted advertising and behavioural profiling, with the opt-out mechanism subject to design standards that prevent "consent fatigue" or friction-based circumvention. Enforcement should sit with a designated regulator empowered to issue binding technical standards.
2 Statutory Technical Advisory Authority
Establish, by law, a permanent body of AI and emerging-technology specialists with a formal seat in the policy formulation chain — analogous to the role played by economic advisory bodies in fiscal policy. This body's mandate should include mandatory technical review of proposed legislation affecting digital markets, AI systems, and data governance prior to introduction, with its assessments placed on public record.
3 Algorithmic Transparency and Audit Framework
Mandate structured disclosure of source code, model architectures, and training methodologologies of large language models and highimpact algorithmic systems to a designated government audit authority, operating under a tiered confidentiality regime to balance transparency with legitimate trade-secret protection. The audit mandate should specifically cover bias and discriminatory outcomes, self-preferencing in ranking or recommendation systems, manipulative interface design ("dark patterns"), and noncompliant consent architecture. Audit findings should carry binding remediation timelines.
4 Cloud and Data Jurisdiction Framework
Codify, in statute, the principle that any data centre or cloud infrastructure, irrespective of physical location, that stores or processes the personal data of Indian citizens falls within Indian jurisdiction for purposes of legal access, data protection enforcement, and national security oversight, in a manner analogous to the extraterritorial jurisdiction recognised over diplomatic premises under international law. This framework should specify the conditions for lawful access requests, mutual legal assistance interoperability, and conflict-of-law resolution mechanisms with foreign jurisdictions.
5 Vicarious Liability for Autonomous Physical AI Systems
Codify a statutory liability regime under which developers and deploying entities of physical AI systems performing detection, decision, and use-of-force functions, including target identification and engagement, such as neutralising hostile drones, bear vicarious liability for harms arising from system operation. This regime should rest on the explicit legal premise that such systems possess no independent legal agency and operate strictly per design, training data, and configuration choices made by identifiable human actors, with liability allocation rules addressing multi-party development chains (model developers, integrators, and end deployers). 7. Conclusion The advent of AI is due to the pain and aspiration of industries for centuries.
Artificial Intelligence, once a distant vision of technological potential, has now materialised as an omnipresent force transforming governance, competition, and human agency. Through control over data pipelines, algorithmic interfaces, and digital infrastructure, a handful of corporations now possess the capacity to influence not only what markets deliver, but how individuals think, act, and vote. It has been demonstrated that AI monopolies pose a compound threat: they distort competition, weaken democratic oversight, and undermine individual autonomy. The concentrated control over AI technologies allows platform giants to engage in algorithmic collusion, self-preferencing, and data-driven manipulation – practices that exploit legal gaps and outpace regulatory reflexes.
As AI systems increasingly mediate public discourse, regulate attention, and automate decision-making, they are fast becoming political actors without accountability. Traditional legal systems, rooted in territorial sovereignty and reactive governance, are proving insufficient in the face of globally distributed cloud infrastructures and borderless algorithms. The crisis is both legal and philosophical: we are witnessing a breakdown of the classical Westphalian model, where data flows freely across jurisdictions while enforcement remains bounded by national borders. This mismatch has resulted in "law without borders, borders without law. " States are now confronted with a choice - either recalibrate their legislative muscle to meet the scale and speed of AI's influence or cede regulatory ground to unaccountable corporate entities.
In response, this paper proposes the Doctrine of Proportionally Smart Governance, a new regulatory ethos grounded in three pillars: capability, proportionality, and foresight. Proportionality here implies more than balance – it demands that regulatory intelligence be scaled to the complexity and impact of AI technologies. Legislatures must institutionalize technological expertise within the policy-making process, establishing standing bodies of AI specialists who inform, review, and shape laws with precision and agility. This cannot be a temporary response but must become a permanent fixture of 21st-century governance. Equally critical is the need for pre-emptive regulation that does not wait for harms to manifest but anticipates them through system-level audits and risk-based scrutiny.
Government access to AI source code and large language models (LLMs) for compliance audits is no longer optional - it is foundational to ensure that code does not quietly erode constitutional values. This must be complemented by legal clarity on vicarious liability in physical AI systems. Furthermore, this research advances the concept of Cloud and Data Jurisdiction, urging a shift in how nations treat data sovereignty. Just as embassies enjoy extraterritorial recognition, data centres storing citizen data abroad must be treated as sovereign extensions, subject to domestic legal control. Only then can nations truly enforce accountability and prevent data colonialism masked as digital convenience. It is also emphasized that regulation should not become a blunt instrument that stifles innovation. On the contrary, smart governance must cultivate competitive pluralism.
AI is not merely a technological force; it is also a constitutional force. It is shaping how laws are interpreted, how power is exercised, and how rights are experienced. If democratic states fail to update their governance frameworks proportionately, they risk being governed not by elected institutions but by proprietary algorithms. Reclaiming digital sovereignty, therefore, is not just a policy challenge. It is a democratic imperative. Governments must assert, with clarity and courage, that while innovation may be driven by code, nations must continue to be governed by constitutions.