India’s Digital Tax Dilemma: - AI and Platform revenues

-by Neeraj Kushawah, Student at Gujarat National Law University, Gandhinagar

Introduction

India’s digital economy is booming - nearly 17 million software developers in India, second only to the United States - yet much of the value created by their activity flows to foreign AI platforms. Advanced Artificial Intelligence (AI) models now routinely surpass human benchmarks, generating tremendous value from Indian users without a local presence.

In practice, firms like OpenAI (ChatGPT), Atrophic (Claude), and Perplexity earn millions of dollars from Indian developers and companies through Application Programming Interface (APIs) and subscriptions. Still, because they have no office, server, or employees in India, they bypass Indian Income Tax entirely.

This “digital Value loop” refers to the cycle where Indian users, developers, or businesses generate value (e.g., via data or paid usage), which foreign AI firms like OpenAI, Anthropic, and Perplexity monetize- yet India cannot tax that income.

India has recognised the need to strengthen its digital tax framework. It was an early mover with an equalisation levy: 6% on online advertising revenue was introduced in 2016 (expanded in 2020 to cover online sales of goods and services).

In 2018, India also inserted a “Significant Economic Presence” (SEP) rule into the Income Tax Act to create a digital permanent establishment, triggering tax where a non-resident earns above Rs. 20 Crore or three lakhs’ users engage with its digital services.

However, these measures have been piecemeal and face limits under international agreements. Indeed, India has agreed to repeal unilateral digital tax like the ad levy as part of the global tax deal, but now India wants to capture value from the digital economy, but the existing laws- built for factories and office businesses- fall short in today's AI-driven world.

In this article, the author analyses the current legal framework for taxing AI-driven companies in India, highlights the existing gaps in the tax regime - particularly concerning digital and platform-based business models - and proposes reforms to align India's approach with evolving international standards.

The Legacy PE Doctrine vs Digital Realities

In traditional tax laws, a country may tax a foreign company’s business profit only if the company has a Permanent Establishment (PE), such as a fixed place of business or a dependent agent. Indian tax laws follow the Organization for Economic Co-operation and Development (OECD) model, where the Income Tax Act (Section 92F) defines a PE as a fixed place of business where the taxpayer conducts the operations wholly or partially. Likewise, section 9(1)(i) of the Income Tax Act effectively requires a physical nexus: under current law, only a business connection involving offices, employees, or equipment in India can be taxed. Explanation 2 to section 9(1)(i) even limits taxability to cases involving a dependent agent habitually concluding contracts in India.

These physical presence requirements now look outdated for digital services. Global software and AI firms can interact with Indian users online without any local office or contract signing agent. Newer business models, such as digital businesses that do not require physical presence or any agent in India, are not covered under Section 9(1)(i). So, India's legacy PE rules were crafted for factories and branches, not for internet platforms.

India has tried to catch up. The 2018 Finance Act inserted Explanation 2A into section 9(1)(i) to define “Significant Economic Presence” (SEP) as also creating a business connection. Under that rule, a large payment now, Rs. 20 Crore per year, or a 3-lakh user base in India would establish a tax nexus. A foreign tech firm with Indian subscribers or transactions above those thresholds would have a “Digital PE”. At the same time, India extended its equalisation levy beyond ads to cover online marketplaces and services.  

Despite these stops, many digital models slip through. Supreme court rulings and treaty definitions mean that routine software licenses or cloud services accessed in India often aren’t taxed as “royalties” or PE. For instance, the Supreme Court held that if a US company licenses software to Indian users, India generally cannot tax it- it is not classified as royalties, no equalisation levy applies to AI. API illustrates that revenue generated from Indian markets accrues and is taxed abroad, leaving India’s tax base untouched. The global API management market was valued at USD 5.42 billion in 2024 and is projected to expand to USD 32.77 billion, reflecting a Compound Annual Growth Rate (CAGR) of around 25%. Within this, the India API management market is expected to surge from USD 42.8 million in 2022 to USD 671.8 million by 2032, growing at an even higher CAGR of 31.76%. Despite India's position as one of the fastest-growing demand centres, revenues linked to API consumption are recognised and taxed in foreign jurisdictions such as the US or Singapore. This mismatch arises because the physical PE doctrine – designed for factories and businesses with offices, employees, and servers- does not map onto the digital economy. As a result, India’s digital demand base contributes significantly to global value creation but yields negligible taxable revenue domestically.

Global Efforts at Digital Taxation

Countries worldwide have grappled with this problem. In 2021, the OECD/G20 inclusive Framework secured a landmark two-pillar agreement on international tax.

Pillar One- (Amount A) aims to reallocate the profits of the largest multinationals' residual profits to market jurisdictions, ensuring taxation where users and consumers are located, even without physical presence.   

Pillar Two- (The Global Anti-Base Erosion (GloBE) Rules) imposed a 15% global minimum tax on significant multinational incomes, preventing “race to the bottom” tax competition.

Over 130 countries joined this plan, reflecting its global reach. In parallel, a multilateral tax treaty (the OECD Model Multilateral Convention) has been drafted and is awaiting ratification by sufficient states. India’s repeal of the 6% equilisation levy and the 2% e-commerce levy align with Pillar One commitments, reflecting confidence that coordinated rules will grant taxing rights on digital sales.

Unilateral (DSTs) reflect frustration with outdated nexus rules. The UK’s 2% DST (2020) raised immediate revenue but sparked trade tensions with the US, while the EU’s shelved DST plan highlights the difficulty of aligning member states while awaiting OECD Pillar One. Countries like France, Italy, and Spain imposed DSTs on ads and intermediation, generating fiscal gains but facing retaliation threats. Over 25 jurisdictions now apply DSTs,  but they risk double taxation, legal disputes, and friction in global trade, underscoring the urgency for a coordinated international solution.

At the same time, more than 60 countries have extended Value Added Tax /Goods and Services Tax to cover digital services consumed within their borders. These measures vary in scope and rate, but all reflect the common goal of taxing values created by digital platforms.  

Lastly, the world is in transition. The OECD/Pillar One vision has wide support, but negotiations, especially on “Amount B rules,” remain unsettled as of 2025. Pillar Two is already being implemented in many OECD countries. India continues to push for user-based nexus rules in OECD and UN discussions, underscoring the need for a domestic digital tax framework even as unilateral levies are scaled back.  

Legal & Practical Challenges

Any digital reform will face hurdles. Enforcement is inherently complex- tracking offshore platforms' profits and user activity demands new data collection and audit capabilities. India would need to mandate disclosure, for example, requiring platforms to report Indian users' revenue, and build tech tools to estimate value generation.

There is also a risk of double taxation. If India taxes cross-border digital income under new rules, the source country may also tax it under existing criteria. Careful treaty negotiation or unilateral relief measures would be needed to avoid unfair piling on of taxes. Likewise, any DST could be challenged as discriminatory under WTO rules unless crafted neutrally. India must work these issues into its design by targeting only foreign-sourced digital income or offering credits for foreign taxes paid.

Business concerns will loom large. The global tech industry argues for stable and predictable rules. For instance, after India moved to repeal its equalisation levy, the US tech lobby praised the decision as restoring a non-discriminatory environment for foreign companies. Indian authorities must balance revenue goals against the risk of discouraging investment. This means engaging in dialogue with industry stakeholders, just as data centre investors have lobbied for clarity on PE rules. It may be wise to phase in changes, issue draft guidelines, and consider safe harbour provisions to give businesses time to adapt.

Concluding, reforms must be legally sound and practically workable. Policymakers should carefully analyse WTO and treaty obligations, possibly seeking advance rulings or amending Double Taxation Avoidance Agreements(DTAs) to accommodate new definitions. Simultaneously, strong stakeholder consultation will be essential to anticipate conflicts; if it is too cautious, it will miss out on tax revenue from a rapidly growing sector. A balanced approach combining clear rules with open dialogue is therefore vital. 

Policy Blueprint for India

To address these challenges, the Author proposed that these existing reforms should be considered from a new perspective. India needs a pragmatic two-pronged strategy that modernises its laws while coordinating internationally.  

Firstly, expand the PE nexus by introducing a “digital PE” test, taxing foreign firms based on digital interactions, user data exploitation, or AI service monetisation, and complement it with a revived equalisation levy or digital services tax as a temporary measure.

Secondly, blend multilateral and unilateral tools by staying active in OECD/United Nations negotiations while strengthening domestic enforcement through transfer pricing, Country by Country Reporting (CbCR), and digital PE audits.

Lastly, use targeted tax incentives such as concessional rates or settlement schemes to encourage voluntary compliance and revenue reporting by foreign firms in India.

Conclusion and the way forward

India’s case of modernising its tax laws is urgent but needs realism. The digital revolution cannot be ignored – Ad revenue is not a niche problem but a systemic one, given the millions of users plugged into global platforms. Yet alarmism will not help; a measured reform agenda is needed. At the legislative level, India should clarify the law by updating Section 9 to explicitly cover significant digital presences or enacting a dedicated digital service tax. At the administrative level, clear guidelines and capacity building will be essential.

Crucial India must also keep up the pace in international diplomacy. It should work within the Organisation for Economic Co-operation and Development (OECD) inclusive framework and UN tax incentives to reshape the global tax regime in line with the realities of platform value creation. This means pushing for user-based allocation rules and ensuring any multilateral agreement allows India's rightful taxing rights.