Buy It And Bury It: The Rise Of Competitive Denial M&A And What It Means For India

Abstract

 This article examines the emergence of "competitive denial M&A" a revolutionary acquisition strategy exemplified by Anthropic’s $300 million acquisition of Stainless in May 2026, where the target company was purchased primarily to deny competitors access to shared infrastructure rather than to achieve traditional synergies. The AI company has transitioned from fundraising startup to aggressive acquirer, fundamentally rewriting M&A playbooks in the artificial intelligence sector. The article also provides comprehensive analysis of Anthropic's unprecedented growth trajectory and the funding mechanics that enable such aggressive acquisition strategies.

The analysis extends to Indian regulatory implications, examining how the Competition Act, 2002 (particularly Section 6's Appreciable Adverse Effect on Competition test, Section 4's abuse of dominant position provisions and Section 5's combination thresholds) and Companies Act, 2013 (Sections 230-232 on mergers requiring NCLT approval and shareholder consent) would apply to competitive denial acquisitions in India. The article investigates whether similar "killer acquisitions" have occurred in India's tech sector, analysing if India's Competition Commission has identified data concentration, compute access, talent hoarding and platform gatekeeping as key AI competition concerns in its 2025 study.

For Indian M&A professionals, founders and investors, the article provides actionable insights on India's position in the AI M&A landscape, including opportunities for Indian AI startups as acquisition targets, risks of regulatory blockage and the strategic implications of global AI M&A trends for the Indian ecosystem. The analysis concludes that while competitive denial M&A represents a revolutionary shift in global AI acquisition strategy, India's regulatory framework is better positioned to prevent anti-competitive acquisitions due to stricter CCI powers, NCLT oversight and upcoming Digital Competition Law, potentially making India's approach the global standard for regulating AI M&A transactions.

Keywords: Competitive Denial M&A, Anthropic, AI Acquisitions, Competition Act 2002, Killer Acquisitions

1.       The Playbook Nobody Saw Coming

There is an old line in competitive strategy - if you cannot beat a rival’s product, make sure they cannot build it. For most of industrial history, that meant poaching talent, buying patents or flooding the market with price cuts. In May 2026, Anthropic did something subtler and arguably far more effective.

On 18 May 2026, Anthropic announced the acquisition of Stainless which is a New York startup founded in 2022 by former Stripe engineer Alex Rattray. Stainless did something unglamorous and utterly essential which was, it automatically converted into an Application Programming Interface (“API”) specification which is production-ready Software Development Kit (“SDK”), across TypeScript, Python, Go, Java, Kotlin and more. Every time a developer wrote a line of code to call an AI API, there was a good chance that the SDK wrapping that call had been generated by Stainless.

The headline looked like a routine developer-tools acquisition. The reality was considerably sharper. Stainless had not merely been building SDKs for Anthropic. It had been building them for OpenAI, Google, Meta’s Llama Stack, Cloudflare, Replicate, Runway, Groq, LangChain and others besides. In other words, a single startup quietly sat beneath the developer infrastructure of almost the entire AI industry.

Then Anthropic bought it. And shut the hosted generator down. Anthropic is switching off the substrate that OpenAI, Google, Cloudflare and others leaned on for everyone except the team that now owns it. This article is about what that actually means for creators at large, not just for Silicon Valley but for Indian Mergers and Acquisitions (“M&A”), Indian competition law and the Indian founders navigating a rapidly changing technology market, that question is no longer theoretical.

2.       The Deal: What Happened and Why It Matters

• The infrastructure nobody noticed

To understand the Stainless transaction, we have to understand what an SDK actually is. Think of AI APIs as the electricity grid. The models are Claude, GPT, Gemini are the power stations. But a power station is useless without the wiring that carries electricity into your home. SDKs are that wiring: the libraries that let a developer in Bengaluru or Berlin write a line of Python and call a frontier AI model without interacting with raw Hypertext Transfer Protocol (“HTTP”).

Maintaining that wiring is tedious, expensive and easy to break. Every time an AI company updates its API, it adds a parameter, changes a response format, deprecates an endpoint. Every official SDK needs to be updated manually across every language. OpenAI reportedly tried managing this internally before deciding the maintenance burden was too high and turning to Stainless. Google did the same. So did most of the industry.

Stainless solved this with automation, by feeding it an API specification and it generated and updated SDKs across six languages simultaneously. By the time of its acquisition, Stainless had already powered every official Anthropic SDK since the earliest days of the Claude API.

2.2.    The acquisition and the shutdown

Anthropic did not disclose the financial terms of the deal. The Information, which first reported that talks were underway, put the price at more than $300 million more than double Stainless’s last private valuation of $150 million from a December 2024 Series A backed by Sequoia Capital and Andreessen Horowitz.

The strategic value, however, lay not in the revenue Stainless was earning but in what happened next. On the day of the announcement, Stainless stated that it would wind down all hosted products, including the SDK generator itself. New signups were immediately closed. Existing customers retained ownership of the SDKs they had already generated and the right to modify them but they lost the automated update pipeline that had been doing the heavy maintenance lifting.

For OpenAI and Google, the impact was immediate and uncomfortable. The tooling that powered their official developer libraries, the thing millions of developers downloaded every week without knowing or caring who built it, now belonged to their most aggressive competitor.

2.3.    Four acquisitions, one direction

The Stainless deal did not happen in isolation. In the six months preceding it, Anthropic had also acquired Bun (the JavaScript runtime, December 2025), Vercept (a computer-use AI startup, February 2026) and Coefficient Bio (a biotech AI company, April 2026, reportedly for approximately $400 million in stock). Each targeted a different layer of what Anthropic is building, the runtime, the autonomous operator, the domain application and now the developer connectivity layer. The sequence matters. The Bun and Vercept acquisitions built internal capability. The Stainless deal removed capability from rivals. That distinction is the difference between vertical integration and competitive denial, a distinction that competition regulators, including India’s Competition Commission of India (“CCI”), are increasingly likely to scrutinise.

3.       Naming the Strategy: What Is Competitive Denial M&A?

The term "killer acquisition" has been in the competition law vocabulary for over a decade. It describes transactions where the primary motive is to eliminate a competitive threat rather than to develop the target’s technology. Competitive denial M&A is a related but distinct concept. Here, the target is not a competitor. It is a shared supplier, a company that sits in the infrastructure layer beneath multiple competitors and provides a service all of them depend on. The acquirer does not buy the supplier to destroy a rival’s product directly. It buys the supplier, integrates the technology exclusively and removes what was previously neutral infrastructure from the shared market. The effect on rivals is not elimination of a competitor but foreclosure of an input.

The distinction matters legally, because most merger control regimes are designed to assess horizontal overlaps (two competitors merging) or vertical integration (a company buying its supplier). Input foreclosure, the theory of harm most relevant to competitive denial M&A is a recognised but less frequently tested in Indian jurisprudence.

4.       The Fable 5 Epilogue: When the Government Plays Acquirer

If the Stainless acquisition showed what a private company can do by buying infrastructure, the Fable 5 episode showed something equally striking: what the state can do by controlling access to it.

On 9 June 2026, Anthropic launched Claude Fable 5 which was its most capable publicly available model, built on the same underlying architecture as Claude Mythos 5, a restricted-access model available only to select partners under Anthropic’s Project Glasswing. Fable 5 was designed to bring frontier-level reasoning, coding and cybersecurity capabilities to the general public, with additional safety guardrails layered on top. The launch lasted three days.

Anthropic stated that the U.S. government had issued an export control directive on 12 June 2026 requiring suspension of access to Fable 5 and Mythos 5 for all foreign nationals, which in practice led Anthropic to disable both models globally. Anthropic indicated that the directive appeared to stem from government concerns regarding a reported method of bypassing Fable 5’s safeguards but publicly disputed the basis and seriousness of that concern, arguing that the issue was narrow in scope, did not amount to a universally replicable vulnerability and could be addressed through its existing monitoring and defence systems. Anthropic also maintained that comparable capabilities could arise in relation to other publicly available frontier models, including OpenAI’s GPT-5.5. The U.S. Commerce Department subsequently lifted the export control directive in full on 30 June 2026, and Anthropic restored global access to Fable 5 from 1 July 2026, alongside broader (though still not universal) access to Mythos 5 for approved U.S. institutions.

The relevance of this episode to M&A and competition law may not be immediately obvious, but it is very real. The Fable 5 suspension illustrates a risk that Indian M&A practitioners and their clients must increasingly factor into transactions involving AI infrastructure, a government acting on security grounds and under no obligation to provide detailed reasons or follow a transparent process, can effectively deny access to a technology at scale overnight. Control over AI infrastructure is no longer just a question of competition. It is also becoming a question of sovereignty.

5.       Has India Seen This Before?

India has not yet seen a transaction that maps onto competitive denial M&A as neatly as the Stainless deal. That said, Indian M&A has produced several transactions that test adjacent ideas and it is useful to examine them carefully rather than forcing them all into a single category.

Reliance–Hathway and Reliance–DEN Networks – Jio’s acquisitions of stakes in these cable distribution networks gave it access to last-mile physical infrastructure, the actual cables and connections reaching into homes, that its rivals in broadband and cable also depended on for distribution. The closer parallel here is control over a shared distribution layer, rather than an outright shutdown of a competitor’s access to it. Reliance didn't switch anything off; it simply gained a structural advantage in owning what others needed to reach the customer.

Zomato–Blinkit - This was primarily a horizontal move into quick commerce, and it would be a stretch to describe it as competitive denial in the strict sense. What it does illustrate well is the CCI’s growing comfort with scrutinising platform acquisitions where the real question isn't today’s overlap in revenue, but tomorrow’s market structure, who ends up controlling the category once the dust settles.

Sun Pharma–Ranbaxy - This is the closest Indian cousin to the pharmaceutical "killer acquisition" literature that gave the whole concept its name. Consolidation involving a concentrated generics pipeline raises the classic concern of a large player absorbing a rival’s product portfolio — closer to eliminating a competitor outright than to denying shared infrastructure, but still useful as the domestic reference point for the underlying idea.

Bharti Airtel–Telenor India - A spectrum and subscriber consolidation in telecom, where an already dominant operator absorbed a smaller rival’s assets, reducing the number of independent players that a customer or a competing operator, could turn to in the affected circles. This edges closer to the Stainless pattern in one respect: it was about acquiring scarce, shared infrastructure (spectrum) rather than simply growing a customer base.

None of these transactions is a perfect analogy for Stainless. What they do have in common, however, is a recurring competition-law question: when does an acquisition merely make the buyer larger and when does it also reduce the options available to everyone else in the market? A Stainless-style transaction, if it arose in India, would force the CCI to confront that question directly in the context of AI infrastructure.

6.       Would Indian Competition Law Actually Catch a Deal Like This?

India’s Competition Act, 2002 provides three broad tools that are relevant to a transaction of this kind.

Combination review which requires acquisitions above certain thresholds to be notified to the CCI before closing. Historically, those thresholds were tied to the assets or turnover of the parties. That posed an obvious problem for a transaction like Stainless. A small developer-tools company may have modest revenue and a limited asset footprint, while still sitting at the centre of a strategically important industry layer. The 2023 amendments addressed this gap by introducing a deal-value threshold. Transactions above a specified value, involving a target with substantial business operations in India, must now be notified even if the target’s traditional financial metrics look modest. This is a significant development for digital and infrastructure acquisitions, because it allows the CCI to examine deals whose strategic importance would otherwise escape review.

The second tool is the CCI’s power to block, or impose conditions on, a combination likely to cause an appreciable adverse effect on competition. In practice, that means looking at barriers to entry, input foreclosure, the effect on suppliers and customers and the likely consequences for innovation and market structure. A transaction that removes shared developer infrastructure from a wider ecosystem would naturally invite this kind of substantive scrutiny.

The third tool becomes relevant after the acquisition if the acquirer holds a dominant position in a relevant market. A dominant enterprise that uses its position to deny market access to rivals or to leverage strength in one market to protect itself in another, may face abuse-of-dominance scrutiny independently of whether the acquisition itself was notified or cleared. In other words, even if the merger review stage is missed or passed, the post-acquisition conduct may still become the subject of competition enforcement.

It is also worth noting that the CCI’s 2025 study on AI and competition specifically identified concerns such as data concentration, compute access, talent hoarding and platform gatekeeping. That suggests the regulator is already alert to the kinds of structural risks that a Stainless-style transaction raises, even if no Indian case has yet squarely tested the point.

7.       The Deal-Value Threshold: Why It Was Built for This Moment

The most important Indian regulatory development for competitive denial M&A is the introduction of the deal-value threshold in 2023. Under the old regime, the conventional Section 5 thresholds would likely not have caught a transaction like Anthropic–Stainless. Stainless was not a large company by revenue and its Indian asset footprint was likely negligible.

But a transaction valued at over $300 million involving a target with meaningful Indian developer adoption and with products used by Indian engineering teams across the AI ecosystem, could now trigger the deal-value threshold if the “substantial business operations in India” requirement is met. The CCI has not yet issued definitive guidance on how it will interpret that standard, but the legislative purpose is fairly clear: to capture strategically significant digital and infrastructure deals that traditional turnover-based thresholds would miss.

For Indian M&A practitioners, the implication is immediate. In transactions involving AI tooling, developer infrastructure, data platforms or compute layers, deal-value analysis can no longer be treated as an afterthought. It has to be part of the regulatory assessment from the beginning.

8.       Where Does India Stand? Opportunities and Risks

• India as an acquisition target market

India’s AI startup ecosystem is maturing quickly. Companies working on foundation models, developer tooling, vertical AI applications and AI-enabled business process automation are increasingly on the radar of global acquirers, including frontier labs that are now moving aggressively in the acquisition market.

For Indian founders and investors, this creates a distinct category of risk. A global acquirer may be interested in an Indian AI company not because of its India revenue or user base, but because of the infrastructure role it plays in a wider developer ecosystem. If that company is acquired and its product is shut down, made exclusive or otherwise removed from neutral circulation, the downstream cost will not be borne only by foreign competitors. Indian developers and Indian enterprises that relied on the same infrastructure will also feel the effect.

This is not a hypothetical concern. India has substantial developer communities using open and neutral infrastructure tools across AI frameworks, API tooling and cloud-native development. If those tools are absorbed into competitive denial transactions without meaningful scrutiny, the Indian ecosystem will bear the foreclosure effect regardless of where the deal was signed.

8.2.    Regulatory risk for inbound acquirers

For global AI companies considering acquisitions of Indian targets or acquisitions of global targets with Indian developer adoption, the 2023 deal-value threshold creates a genuine filing risk that did not previously exist. A failure to notify a notifiable combination is a serious CCI enforcement matter, with penalties potentially reaching one percent of global combined assets or turnover.

There is also substantive risk. The CCI now has broader powers to impose remedies on transactions that raise competition concerns, including divestitures, behavioural conditions and access-related obligations. In theory, an acquirer that shuts down a competitor-facing product after closing could face an order requiring it to restore access or divest the acquired asset.

8.3.    The Digital Competition Law horizon

India is also in the process of developing a Digital Competition Law, modelled in part on the EU’s Digital Markets Act. If enacted in its proposed form, it would impose ex ante obligations on systemically significant digital enterprises, including restrictions on self-preferencing, obligations relating to data use and curbs on acquisitions that entrench existing market power. A Stainless-style transaction, where a dominant AI platform acquires shared developer infrastructure and makes it exclusive, is very close to the kind of conduct such a framework appears designed to address.

India’s evolving competition framework may, in fact, be better positioned than that of many peer jurisdictions to address competitive denial M&A in real time. The combination of the deal-value threshold, the substantive AAEC framework and a forthcoming digital competition regime gives the CCI a potentially robust toolkit for dealing with AI infrastructure foreclosure.

8.4.    Conclusion: India’s Moment to Set the Standard

The Anthropic–Stainless transaction and the Fable 5 suspension arrived through entirely different mechanisms, one commercial and one regulatory, but they point toward the same underlying shift. The AI industry has now reached a stage where it is not enough to have the best AI model; controlling the infrastructure behind those models has also become equally important in deciding who gets ahead in the market. They are acquiring the pipes through which those models are built, distributed and used.

For India, that creates both a challenge and an opportunity. The challenge is obvious: Indian developers, Indian enterprises and Indian startups are downstream users of global AI infrastructure and competitive denial in that infrastructure will affect them directly. The opportunity is equally significant. India’s regulatory framework, with its deal-value threshold, its competition-law tools around foreclosure and market access and its proposed digital competition regime, is reasonably well placed to respond. The more difficult question is whether the CCI will move quickly enough and with sufficient conceptual clarity, to apply that framework before the structure of AI infrastructure markets changes in ways that are hard to reverse. Stainless was wound down in a matter of days. Competition review moves much more slowly. India has a real opportunity to become the first jurisdiction to develop a coherent legal response to competitive denial M&A in AI. If it does, its approach will matter far beyond Indian borders.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.