Merger Control And AI Start‐Ups - Adequacy Of Existing Thresholds
Author: LawSphereX
Category: Corporate & Insolvency
Published:
Updated:
MERGER CONTROL AND AI START‐UPS: ADEQUACY OF
EXISTING THRESHOLDS
Sandhiya S, LLM (Corporate Law), SRM School of Law, SRM Institute of Science and Technology, Kattankulathur
ABSTRACT
This research investigates whether the current merger notification thresholds and provisions are sufficient to effectively target the impact of the competitive dynamics of the artificial intelligence (AI) sector, specifically so-called “killer acquisitions” of developing AI startups by established technology companies. The research will focus on the legislative reform under the Competition (Amendment) Act, 2023 and the notified Combination Regulations and notifications (including the Deal-Value Thresholds framework), the jurisprudence and best practice in the main jurisdictions (EU, US), and the literature on “killer acquisitions,” market tipping, and developing competitor theories of harm.
This research will integrate a doctrinal analysis of the law, an assessment of the policy, and a small-scale empirical analysis of high-profile AI-related acquisitions to assess whether the current thresholds (asset/turnover and DVT) are sufficient to effectively target acquisitions that may foreclose future innovation or concentration in the AI layer. The final section of the paper will offer recommendations on threshold design, information collection, and remedial policies that the competition authority might pursue to reconcile the innovation gains from acquisitions with competition protection.
LIST OF ABBREVIATIONS
S. No. Abbreviation
Full Form
1. AI
Artificial Intelligence
2. API
Application Programming Interface
3. CCI
Competition Commission of India
4. CUP
Cambridge University Press
5. DG Competition Directorate-General for Competition (European Commission)
6. DOJ
Department of Justice (United States)
7. DVT
Deal-Value Threshold
8. EU
European Union
9. EUMR
European Union Merger Regulation
10. FTC
Federal Trade Commission (United States)
11. IP
Intellectual Property
12. ICRIER
Indian Council for Research on International Economic Relations
13. MCA
Ministry of Corporate Affairs
14. M&A
Mergers and Acquisitions
15. OECD
Organisation for Economic Co-operation and Development
16. R&D
Research and Development
17. SSRN
Social Science Research Network
18. US
United States
CHAPTER I
Introduction
1.1 Background of the Study
Merger control constitutes a central pillar of competition law, aimed at preventing market structures that may substantially lessen competition or adversely affect consumer welfare. In India, the regulation of mergers and acquisitions is governed by the Competition Act, 2002, which empowers the Competition Commission of India (CCI) to scrutinise combinations that exceed specified jurisdictional thresholds based on assets and turnover.1 Traditionally, these thresholds were designed to filter transactions unlikely to raise competition concerns while ensuring administrative efficiency.
1 Competition Act, 2002 (India), ss 5–6.
However, rapid technological change and the growth of digital markets—particularly artificial intelligence (AI)—have challenged the adequacy of conventional merger control frameworks. AI startups often possess highly competitive and strategic value derived from intellectual property, data, algorithms, and specialised human capital rather than current revenues or physical assets. 2 As a result, acquisitions involving AI startups frequently fall below traditional turnover-based thresholds, despite their potential to reshape future market competition. Globally, competition authorities and scholars have expressed concern that dominant firms may acquire such startups to eliminate future competitive threats, a phenomenon commonly described as “killer acquisitions.
”3 These acquisitions may not produce immediate price effects but can significantly reduce innovation, diversity, and long-term consumer welfare. Recognising these concerns, jurisdictions such as the European Union and Germany have experimented with alternative jurisdictional tools, including transaction value thresholds and discretionary referral mechanisms. 4 In India, these global developments culminated in the enactment of the Competition (Amendment) Act, 2023, which introduced significant reforms to merger control, including the statutory basis for a Deal-Value Threshold (DVT). 5 This reform reflects an acknowledgment that innovation-driven markets require more flexible and forward-looking regulatory tools. Against this background, the present study examines whether India’s revised merger control thresholds are adequate to address acquisitions involving AI startups.
1.2 Statement of the Problem
The central problem addressed in this study is the mismatch between traditional merger notification thresholds and the economic realities of AI-driven markets. While AI startups may exert substantial competitive pressure through innovation and technological potential, they often generate minimal turnover at the time of acquisition. Consequently, their acquisition by dominant firms may escape regulatory scrutiny under conventional asset- and turnover-based thresholds.6
2 P Malik, AI Markets and Competition in India (ICRIER 2024).
3 C Cunningham, F Ederer and S Ma, ‘Killer Acquisitions’ (2018) SSRN.
4 European Commission, Guidance on Article 22 Referrals (2021).
5 Competition (Amendment) Act, 2023 (India).
6 M Ivaldi et al, ‘Killer Acquisitions: Evidence from European Merger Control’ (2020).
Although the introduction of the Deal-Value Threshold seeks to address this gap, questions remain regarding its calibration, scope, and effectiveness in capturing strategically significant AI acquisitions. Furthermore, the assessment of such transactions presents unique challenges relating to prediction of future competition, valuation of intangible assets, and technical complexity. This study therefore investigates whether existing thresholds and analytical tools under Indian competition law are sufficient to prevent anticompetitive consolidation in AI markets.
1.3 Significance of the Study
This research is significant for several reasons. First, it contributes to the evolving discourse on innovation-centred merger control, an area of growing importance in corporate and competition law. Second, it offers a focused analysis of AI startups, an area that has received limited doctrinal attention in Indian legal scholarship. Third, the study has practical relevance for regulators, policymakers, corporate lawyers, and investors engaged in technology-driven mergers and acquisitions. By evaluating the adequacy of current thresholds and drawing on comparative jurisprudence, this research aims to inform future regulatory reform and enhance the effectiveness of merger control in safeguarding competitive markets without unduly stifling innovation.7
1.4 Scope and Limitations of the Study
The scope of this study is limited to merger control under competition law, with specific reference to acquisitions involving AI startups. The primary jurisdiction of analysis is India, with comparative references to the European Union and the United States to derive doctrinal and policy insights. The study focuses on jurisdictional thresholds, notification requirements, and analytical challenges rather than the entire spectrum of competition law enforcement. The study is doctrinal and analytical in nature and does not undertake a comprehensive empirical or econometric analysis of AI acquisitions. Additionally, given the rapidly evolving nature of AI technology and regulatory responses, the analysis reflects the legal and policy position up to 2025.8
7 OECD, Startups, Killer Acquisitions and Merger Control (2021).
8 Competition Commission of India, Market Study on Digital Markets (2023).
1.5 Review of Literature
Scholars and policy agencies have devoted increasing attention to killer acquisitions and the limitations of turnover-based merger notification thresholds in innovation-driven markets. Cunningham, Ederer and Ma’s empirical work demonstrates that incumbents in technologyintensive industries often acquire startups to discontinue competing innovation, thereby harming future competition. 9 Ivaldi and colleagues extend this analysis to European merger control and show that such acquisitions frequently escape scrutiny because the targets generate little or no turnover. 10 Their findings have significantly influenced policy debates on the need for alternative jurisdictional thresholds.
Complementing this scholarship, the OECD and the European Commission have emphasised the importance of flexible merger control tools, enhanced information gathering, and innovation-based assessments in digital markets. 11 Recent law-and-technology studies further highlight the AI stack—comprising infrastructure, foundational models, data, and applications—as a locus of concentrated power where acquisitions can have systemic and cross-market effects. 12 These insights collectively underscore the need to reassess merger control thresholds in the context of AI startups.
1.6 Objectives of the Study
The objectives of this study are:
1. To examine the existing merger control framework under Indian competition law with
respect to AI startup acquisitions.
2. To analyse the adequacy of traditional and deal-value thresholds in capturing
anticompetitive AI-related mergers.
3. To study comparative approaches adopted by the EU and the US in addressing nascent
competition and killer acquisitions.
9 Cunningham, Ederer and Ma (n 3)
10 Ivaldi et al (n 6).
11 OECD, Big Data and Competition Policy (2020).
12 European Commission, Competition Policy Brief: AI and Competition (2024).
4. To identify legal and policy gaps in the current Indian framework and propose
recommendations for reform.
1.7 Research Questions
This study seeks to answer the following research questions:
1. Are existing merger notification thresholds under Indian competition law adequate to
capture acquisitions involving AI startups?
2. To what extent does the Deal-Value Threshold address the problem of killer
acquisitions in AI markets?
3. What lessons can India draw from comparative merger control practices in the EU and
the US?
1.8 Hypothesis
The study proceeds on the hypothesis that while the introduction of the Deal-Value Threshold represents a positive development, existing merger control thresholds under Indian competition law remain insufficient to fully address the competitive risks posed by AI startup acquisitions, particularly those involving nascent competition and innovation suppression.13
1.9 Research Methodology
This research adopts a doctrinal and analytical methodology. Primary sources include statutes, rules, notifications, and decisions of the Competition Commission of India. Secondary sources include academic literature, policy reports of international organisations such as the OECD, and comparative materials from the EU and US. The study employs comparative analysis to evaluate alternative regulatory approaches and synthesises these insights to propose contextspecific reforms for India.14 13 H Hovenkamp and C Shapiro, ‘Horizontal Mergers, Market Structure and Innovation’ (2018) Antitrust Law Journal.
14 M Motta and M Peitz, Competition Policy: Theory and Practice (CUP 2021).
CHAPTER II
Artificial Intelligence Startups, Market Structure and Competition Law Concerns
2.1 Understanding AI Startups and the AI Market Structure
Artificial Intelligence (AI) markets differ fundamentally from traditional product markets due to their reliance on intangible assets such as data, algorithms, computing infrastructure and highly specialised human capital. AI startups typically operate within a layered ecosystem, commonly referred to as the AI value chain or AI stack, comprising: (i) hardware and compute infrastructure, (ii) foundational or general-purpose AI models, (iii) data acquisition and training layers, and (iv) downstream application and service layers. AI startups often enter at the model-development or application layer, focusing on innovation rather than immediate commercialisation. Consequently, their current turnover and asset base may be minimal, even though their strategic competitive significance is high.
Their valuation is driven by expected future innovation, access to proprietary datasets, trained models and specialised engineering talent rather than existing revenues. 15 This structural characteristic creates a regulatory blind spot in traditional merger control regimes that rely primarily on assets and turnover thresholds. In many AI-related acquisitions, the target firm does not satisfy notification thresholds despite possessing the potential to evolve into a strong competitive constraint on incumbents. This is particularly relevant in India, where AI startups are rapidly growing in number but remain revenue-light in their early stages. 16
2.2 Characteristics of AI Startups Relevant to Merger Control
AI startups exhibit several features that complicate competition assessment:
15 P Malik, AI Markets and Competition in India (ICRIER 2024).
16 Competition Commission of India, Market Study on Artificial Intelligence and Competition (2024).
a. Low turnover, high valuation paradox AI startups frequently receive high valuations based on projected future applications rather than current market presence. Venture capital investments in AI reflect expectations of scalability and dominance rather than present sales. As a result, traditional turnover-based thresholds fail to capture acquisitions of firms that may later emerge as disruptive competitors. 17 b. Dependence on data and talent rather than physical assets Unlike manufacturing or service enterprises, AI firms rely heavily on non-rivalrous inputs such as datasets and human capital. Acquisitions may therefore involve the transfer of strategic assets not reflected on balance sheets. Such “acqui-hire” transactions often escape notification requirements. 18 c. Network effects and tipping risks AI markets tend to exhibit strong indirect network effects.
Access to larger datasets improves model performance, which attracts more users, thereby reinforcing market power. Acquisitions that consolidate datasets or restrict access to AI infrastructure can accelerate market tipping and entrench dominance. 19 These characteristics demonstrate why AI markets are particularly susceptible to anticompetitive consolidation through under-the-radar acquisitions, making threshold adequacy a central concern.
2.3 Nascent Competition and the Concept of “Killer Acquisitions”
The concept of nascent competition refers to firms that are not yet direct competitors in a defined market but possess the capability and incentives to become competitive in the future. Traditional merger analysis focuses on current overlaps; however, in innovation-driven markets such as AI, the loss of future competition can be equally harmful. A “killer acquisition” occurs when an incumbent firm acquires a nascent competitor with the
17 M Ivaldi et al, ‘Killer Acquisitions’ (2020) 129 Journal of Political Economy 649.
18 OECD, Startups, Killer Acquisitions and Merger Control (2021).
19 European Commission, Competition Policy Brief: AI and Competition (2024).
objective or effect of discontinuing, delaying or redirecting the target’s innovation pipeline, thereby preserving its own market position. Empirical studies have documented such acquisitions particularly in technology-intensive sectors.20 In AI markets, killer acquisitions may manifest in several forms: • termination of independent AI model development post-acquisition, • integration of the startup’s technology solely to improve the acquirer’s core product, • restriction of access to datasets or APIs previously available to third parties, and • absorption of key personnel, effectively neutralising future rivalry. These outcomes may not immediately raise prices or reduce output but can significantly diminish dynamic competition, innovation diversity and long-term consumer welfare.
2.4 Why Existing Merger Thresholds Fail to Capture AI Acquisitions
Traditional merger control thresholds, including those under the Competition Act, 2002, are primarily based on quantitative financial metrics such as assets and turnover. While administratively efficient, such metrics are poorly aligned with the realities of AI markets. a. Turnover thresholds and innovation markets AI startups frequently operate in pre-commercial stages. Their economic value lies in intellectual property, future applications and research potential. Consequently, acquisitions with profound competitive implications may fall below notification thresholds. This issue was recognised internationally in high-profile technology acquisitions, such as Facebook/Instagram and Google/DeepMind, which were cleared without extensive scrutiny due to low target turnover. 21 b.
Asset based thresholds and intangible value Accounting rules often fail to reflect the true competitive significance of datasets, trained
20 C Cunningham, F Ederer and S Ma, ‘Killer Acquisitions’ (2018) SSRN.
21 Lina Khan, ‘Amazon’s Antitrust Paradox’ (2017) 126 Yale Law Journal 710.
models or proprietary algorithms. As a result, asset-based thresholds underestimate the market power implications of acquiring AI startups. c. Deal structuring and regulatory arbitrage Acquisitions may be structured as minority investments, staged acquisitions or asset purchases to avoid triggering notification thresholds. In AI markets, where incremental acquisitions of talent or data can cumulatively lead to dominance, such structuring poses serious enforcement challenges.22
2.5 The Deal-Value Threshold (DVT): Partial Solution or Structural Fix?
The introduction of the Deal-Value Threshold (DVT) under the Competition (Amendment) Act, 2023 represents a significant shift in Indian merger control. By requiring notification of transactions exceeding a specified deal value where the target has substantial business operations in India, the DVT seeks to address the limitations of turnover-based thresholds.23 While the DVT is a welcome reform, its effectiveness in capturing AI acquisitions remains constrained by several factors:
1. High threshold level - AI acquisitions involving early-stage startups may still fall
below the notified deal-value amount despite strategic importance.
2. Business presence requirement - startups with R&D or user bases in India but limited
commercial operations may not satisfy this criterion.
3. Valuation uncertainty - earn-outs, stock-based consideration and internal
restructurings complicate the determination of deal value. Therefore, although the DVT expands regulatory reach, it does not fully resolve the structural mismatch between merger thresholds and innovation-driven competition risks.
2.6 Innovation Harm and Consumer Welfare in AI Markets
Competition law traditionally focuses on price effects and output restrictions. However, AI
22 FTC, Partnerships Between Cloud Service Providers and AI Developers (2024).
23 Competition (Amendment) Act, 2023; Competition (Criteria of Combination) Rules, 2024.
markets necessitate a broader understanding of consumer welfare that includes: • reduced innovation incentives, • diminished quality and diversity of AI applications, and • increased dependency on a few dominant AI providers. The acquisition of AI startups by dominant firms may create innovation bottlenecks, limiting alternative technological trajectories. Such harms are difficult to quantify but may have longlasting implications for economic growth and technological sovereignty, particularly in emerging economies such as India.24 Recognising innovation harm as a legitimate competition concern is therefore essential for effective merger control in AI markets.
2.7 Relevance of AI Market Dynamics for Indian Competition Law
India’s AI ecosystem is characterised by a large number of early-stage startups, increasing foreign investment and growing integration with global technology platforms. The Competition Commission of India has acknowledged the need to adapt competition tools to digital and AI-driven markets through market studies and policy engagement. However, without further refinement of thresholds, sector-specific triggers or enhanced information-gathering powers, the existing framework risks allowing significant AI-related acquisitions to proceed unchecked. This could entrench market power, reduce domestic innovation capacity and undermine the objectives of the Competition Act.
CHAPTER III
Comparative Case Law and Regulatory Practice on AI and Digital-Era Mergers
3.1 Rationale for Comparative Analysis
Merger control in innovation-driven markets such as artificial intelligence presents challenges that transcend national boundaries. Since Indian jurisprudence on AI-specific acquisitions
24 OECD, Artificial Intelligence, Innovation and Competition (2023).
remains limited, comparative analysis of enforcement practices in the European Union and the United States provides valuable doctrinal guidance. These jurisdictions have confronted similar issues relating to nascent competition, below-threshold acquisitions and innovation harm and their evolving approaches offer lessons for India’s developing framework.
3.2 European Union: Innovation Competition and Below-Threshold Acquisitions
3.2.1 EU merger control framework and innovation theory
The European Union has progressively shifted from a static price-effects analysis towards a more dynamic understanding of competition that includes innovation and potential competition. Under the EU Merger Regulation (EUMR), mergers are assessed for their effects on effective competition, including future market structure and innovation incentives.25 The European Commission has explicitly recognised that in digital and AI markets, small firms may exert significant competitive pressure through innovation, even if they lack current market share. Consequently, the EU has developed doctrinal tools to scrutinise acquisitions that eliminate potential competitors.
3.2.2 Facebook/WhatsApp and Facebook/Instagram
The acquisitions of Instagram (2012) and WhatsApp (2014) by Facebook (now Meta) are frequently cited as archetypal cases illustrating the limitations of turnover-based thresholds. At the time of acquisition, both targets had minimal revenues and were cleared without detailed innovation-based analysis. Subsequent regulatory reflection suggests that these acquisitions significantly altered the competitive landscape of social networking and messaging markets. The European Commission later acknowledged that the loss of potential competition should have been given greater weight.26 These cases have become central to the global debate on whether merger control should intervene earlier in digital markets to prevent concentration through serial
25 Council Regulation (EC) No 139/2004 (EU Merger Regulation).
26 European Commission, Competition Policy Brief on Digital Markets (2021).
acquisitions.
3.2.3 Google/DeepMind and AI-specific concerns
Google’s acquisition of DeepMind Technologies represents a critical example for AI markets. DeepMind was a research-oriented AI company with limited commercial turnover but immense technological potential. The transaction was not subject to detailed merger scrutiny due to jurisdictional thresholds. Policy commentators argue that the acquisition illustrates how control over foundational AI research and talent can shape long-term market structure. The European Commission has since emphasised the need to assess acquisitions involving foundational AI models more closely, even where current revenues are negligible.27
3.2.4 Article 22 referrals and expanded jurisdiction
A significant EU development is the revitalisation of Article 22 EUMR referrals, allowing
Member States to refer mergers that do not meet EU thresholds but may affect competition. This mechanism has been used to assert jurisdiction over potentially anticompetitive digital acquisitions. The policy shift demonstrates an acknowledgment that traditional thresholds are insufficient in innovation markets. This approach is particularly relevant for AI acquisitions, where early intervention may be necessary to preserve competitive market structures.28
3.3 United States: Nascent Competition and Antitrust Revival
3.3.1 US antitrust approach to killer acquisitions
US antitrust authorities, particularly the Federal Trade Commission (FTC) and the Department of Justice (DOJ), have increasingly focused on nascent competition and serial acquisitions by dominant technology firms. While US merger law historically emphasised price effects, recent enforcement signals a broader concern with innovation suppression. Academic scholarship, notably the work on “killer acquisitions,” has influenced US
27 European Commission, Competition Policy Brief: AI and Competition (2024).
28 European Commission, Guidance on Article 22 Referrals (2021).
enforcement thinking by providing empirical evidence that incumbents acquire startups to discontinue competing innovation.29
3.3.2 FTC scrutiny of AI partnerships and acquisitions
The FTC has expressed concern over partnerships and acquisitions involving AI developers and large cloud service providers. In its 2024 report on AI partnerships, the FTC warned that vertical integration between AI startups and dominant infrastructure providers could lead to foreclosure, reduced interoperability and innovation bottlenecks.30 The US approach increasingly relies on information-intensive investigations, extensive document requests and forward-looking market assessments tools that go beyond simple threshold tests.
3.3.3 Microsoft/OpenAI and structural concerns
Although structured as a strategic partnership rather than a full acquisition, the Microsoft/OpenAI relationship has drawn regulatory attention globally. Competition authorities have examined whether such arrangements effectively confer control over critical AI capabilities, raising questions about the adequacy of formal ownership thresholds. This case illustrates that control in AI markets may arise through contractual arrangements, not just share acquisitions, a lesson of direct relevance to Indian merger control.
3.4 Indian Merger Control Practice and AI Markets
3.4.1 Evolution of Indian merger control
Indian merger control under the Competition Act, 2002 traditionally relied on asset and turnover thresholds to determine notifiability. While effective in conventional markets, this approach struggled with digital and innovation-driven sectors. The Competition (Amendment) Act, 2023 marks a turning point by introducing the Deal- Value Threshold (DVT) and strengthening the Competition Commission of India’s (CCI)
29 C Cunningham, F Ederer and S Ma, ‘Killer Acquisitions’ (2018) SSRN.
30 Federal Trade Commission, Partnerships Between Cloud Service Providers and AI Developers (2024)
enforcement toolkit.31
3.4.2 CCI engagement with digital and AI markets
The CCI has undertaken market studies on digital markets and publicly acknowledged the competition challenges posed by data concentration and AI technologies. However, as of 2025, there is limited published decisional practice specifically addressing AI startup acquisitions. This absence reflects both the novelty of the sector and the historical constraints imposed by jurisdictional thresholds. The new statutory framework creates opportunities for future enforcement, but its success will depend on how actively the CCI applies its expanded powers.
CHAPTER IV
Analysis, Findings and Major Outcomes
4.1 Assessment of Existing Merger Control Thresholds for AI Startups
The central research question of this study concerns whether existing merger control thresholds in India are adequate to capture anticompetitive acquisitions involving AI startups. The analysis reveals a partial but incomplete alignment between the legal framework and the economic realities of AI markets. The introduction of the Deal-Value Threshold (DVT) represents a substantial improvement over purely turnover-based criterion. It acknowledges that valuation can be a proxy for competitive significance in innovation-driven sectors. However, the high threshold level and additional requirements relating to business presence in India limit its effectiveness in capturing early-stage AI acquisitions.
4.2 Structural Gaps in Threshold Design
Several structural gaps persist: a. Early - stage acquisitions remain under captured Many AI startups are acquired before achieving significant valuation milestones. Such
31 Competition (Amendment) Act, 2023 (India).
acquisitions may strategically neutralise future competitors without triggering notification obligations. b. Focus on formal transactions Merger control thresholds focus on formal acquisitions but may fail to capture control exercised through long-term partnerships, exclusive cloud arrangements or minority investments. c. Static metrics in dynamic markets Thresholds based on fixed numerical values struggle to account for rapidly evolving innovation markets where competitive impact is uncertain but potentially significant.
4.3 Enforcement and Evidentiary Challenges for the CCI
Even where transactions are notified, assessing AI mergers poses substantial challenges: • Counterfactual uncertainty: Determining what the startup would have become absent the acquisition is inherently speculative. • Technical complexity: Evaluating AI models, datasets and training processes requires specialised expertise. • Information asymmetry: Acquiring firms possess superior knowledge of integration plans and strategic intent. These challenges underscore the need for procedural innovations alongside threshold reform.
4.4 Major Findings of the Research
The study yields the following major findings:
1. Traditional merger thresholds are insufficient for AI markets due to the disconnect
between turnover and competitive significance.
2. The Deal-Value Threshold improves detection but does not fully address killer
acquisitions or nascent competition concerns.
3. Comparative jurisdictions increasingly rely on discretionary and innovation-focused
tools rather than rigid thresholds.
4. Indian merger control requires complementary measures to effectively protect
competition and innovation in AI markets.
4.5 Policy and Regulatory Outcomes
Based on the analysis, the following outcomes emerge: • Merger control in AI markets must prioritise innovation preservation over short-term price effects. • Thresholds should serve as screening mechanisms, not exclusive jurisdictional barriers. • Regulatory capacity must evolve to incorporate technical expertise and dynamic assessment tools.
4.6 Implications for Corporate Law and M&A Practice
For corporate law and M&A practitioners, the evolving merger control landscape implies: • increased compliance obligations for AI-related transactions; • heightened due diligence regarding competition risks; • greater regulatory uncertainty in deal structuring. For startups, stricter scrutiny may affect exit strategies but could also preserve competitive opportunities by preventing premature absorption by dominant firms.
CHAPTER V
Case law analysis
A. Google / DeepMind
Although Google’s acquisition of DeepMind (2014) pre-dates recent regulatory focus, it is
often cited by commentators as an example of an acquisition that had large innovation consequences despite limited initial revenues from the target. The concern is that a turnoverbased regime would not have captured the acquisition had it been subject to a pure turnover test at the time. Scholarly commentary and policy papers point to DeepMind as a canonical example of why classical thresholds can fail in dynamic tech markets. B. Facebook / WhatsApp and Instagram
Acquisitions of Instagram (2012) and WhatsApp (2014) by Facebook raised later policy questions about market power and the elimination of nascent rivals. Regulators have reflected post-hoc that earlier scrutiny might have been warranted. These cases illustrate the difficulty of ex ante identification and the importance of forward-looking analysis of potential competition. C. EU & US enforcement on tech acquisitions
The EU and US authorities have started to tackle below-threshold or problematic deals, including reviewing transactions, issuing guidance and amending merger review practice to consider innovation competition. The European Commission has highlighted investment and partnership patterns in generative AI as areas of active interest. The FTC in the US has also signalled greater scrutiny of serial tech acquisitions and strategic partnerships. These authorities’ evolving doctrines on killer acquisitions provide models for India, particularly in the use of expanded evidence collection and innovation-centric analysis. D. Indian practice and CCI engagement with AI
The CCI has issued market studies and guidance signalling attention to AI and digital markets; however, as of the materials reviewed up to mid-2025, there is limited published enforcement decision-level jurisprudence on AI startup acquisitions specifically. The new statutory regime and Combination Regulations broaden the Commission’s tools and suggest increased future activity. For immediate enforcement to be effective, the Commission will need to adopt specialized internal capabilities and possibly sectoral notification tools.
CHAPTER VI
Conclusion & Recommendations
Conclusion The Competition (Amendment) Act, 2023 and the supporting Combination Regulations with DVT represent a meaningful modernization of India’s merger control architecture. They materially improve the Commission’s jurisdictional reach for transactions where deal valuations are significant relative to target turnover. However, DVT calibration, exemptions, and the structural features of many AI acquisitions (acqui-hires, asset purchases, earn-outs or acquisitions for talent/data) mean that important classes of anticompetitive acquisitions may still escape ex ante review. The regulator faces information, technical and evidentiary challenges in assessing innovation harms, requiring supplementary procedural and substantive tools. Recommendations
1. Lower or tier DVT for AI-critical transactions — consider a lower DVT where the
target’s primary business relates to foundation models, critical datasets, or tier-1 AI infrastructure.
2. Sectoral notification mechanism — enable mandatory pre-merger notice for
acquisitions involving clearly-defined AI assets, irrespective of DVT or turnover.
3. Technical disclosure template — require parties to provide a structured technical
appendix in deals involving AI (models, datasets, personnel movement, integration plans).
4. Expert advisory panel — institutionalize access to independent AI technical experts
to aid the CCI in rapid, robust assessment. 5. Interim enforcement powers — strengthen tools for preserving assets (firewalls, nondiscrimination undertakings) pending substantive review.
6. Cross-border cooperation — establish formal channels for information sharing with
EU/US authorities on cross-border AI acquisitions.
7. Public market study & guidance — CCI should publish sector-specific guidance for
M&A in AI to help parties self-assess filing obligations and share best practices.
BIBLIOGRAPHY
1. Competition (Amendment) Act, 2023 (India).
2. Ministry of Corporate Affairs, ‘Notification regarding The Competition (Criteria of
Combination) Rules, 2024’ (G.S.R., 07/03/2024).
3. Competition Commission of India, ‘Filing of Combination Notices — Introduction’
(CCI website).
4. AZB & Partners, ‘Deal Value Threshold (DVT)’ explainer (2024).
5. MergerFilers, ‘India guide’ (updated 2025).
6. European Commission, ‘Competition Policy Brief — AI & Competition’ (Sep 2024).
7. Ivaldi M et al., ‘Killer Acquisitions: Evidence from European Merger Cases’ (2024).
8. FTC, ‘Partnerships Between Cloud Service Providers and AI Developers’ (report).
9. P. Malik, ‘AI Markets and Competition in India’ (ICRIER report)
10. Cunningham C, Ederer F and Ma S, ‘Killer Acquisitions’ (2018) SSRN.
11. Motta M and Peitz M, Competition Policy: Theory and Practice (CUP 2021).
12. Hovenkamp H and Shapiro C, ‘Horizontal Mergers, Market Structure and Innovation’
(2018) 65 Antitrust Law Journal.
13. OECD, Startups, Killer Acquisitions and Merger Control (2021).
14. OECD, Big Data and Competition Policy (2020).
15. Varian H, ‘Artificial Intelligence, Economics and Industrial Organization’ (2023).