In 1602, the Dutch East India Company received a charter from the States-General of the Netherlands to trade spices across Asia. Within fifty years, it had not merely traded. It had built the ports, staffed the courts, written the contracts, and set the prices. The locals who had the spices never controlled what those spices were worth. Three centuries later, when British telegraph cables connected Calcutta to London, India’s economy moved on communication infrastructure that India neither built nor owned. The traffic went where the cable owner decided. The intelligence moved in one direction. The money moved in the other.
The pattern is not new. It has simply found new materials to work with. In the 1990s, India became the world’s software services capital, writing code for foreign products on foreign platforms under foreign IP agreements. India got the salaries. The owners got the companies. Today, with 27 million Indian developers generating more coding data than almost any other country on earth, the same structural offer has arrived again, dressed in the language of democratisation, localisation, and affordable access.
In the week of June 16, 2026, SpaceX announced it would acquire Cursor for $60 billion, folding the AI coding assistant into the xAI division alongside Grok and the Colossus supercomputer cluster. Three days later, Cursor (an American AI software and coding agent) unveiled its India-first localised pricing at Rs 649 per month, roughly a third of its global $20 price, making India the first large market where a major AI coding platform experiments with a stripped-down product at scale. India has more than 27 million developers on GitHub’s platform, second only to the United States, with more than two million joining in 2026 alone.
If you ask me whether the Rs 649 per month Cursor(which is an American software company) Start plan for Indian developers is a good deal, the short answer is: for the individual developer, probably yes. For India as a strategic actor in the AI age, it is one of the more sophisticated traps the country has walked into recently, and most of the commentary around it has completely missed the defence and security mechanisms focussed specifically on market and leaving the strategic contours and angles. The way I read this is that : Cursor’s India push is not generosity. It is a data acquisition strategy dressed as a pricing concession. SpaceX said in its IPO filing that Cursor’s access to developer behaviour is a goldmine that can be used to train the next generation of models. India’s developers are the goldmine. The Rs 649 plan is the mining equipment.
The Starlink Connection Nobody Is Naming Clearly
This is where the argument gets important and where most commentary has stopped short. The Cursor-SpaceX acquisition is not just a developer tools story. It is a stack convergence story, and the stack it is converging with already raises serious security concerns in New Delhi. SpaceX controls Starlink. New Delhi has already frozen or delayed approvals for Starlink’s commercial rollout over concerns about Indian control over a US-based communications operator in wartime or geopolitical tension. Starlink terminals have been found smuggled in Manipur and the Andaman and Nicobar Islands. The Kutniti Foundation has described Starlink explicitly as a technology of geopolitical control tied closely to the US intelligence-military complex.
The interconnectedness that this creates is the insight that matters. If Indian developers build defence-adjacent applications, logistics systems for PSUs, cyber-security tools, or dual-use autonomous systems using Cursor’s agentic workflows, and those workflows run on xAI’s infrastructure, which is integrated with SpaceX’s satellite communications network, India’s ability to audit, re-platform, or isolate those systems under pressure is inherently constrained.
The same entity controls the AI coding environment, the model, and the communications infrastructure beneath it. That is not a technology product. That is a stack-level dependency, and stack-level dependencies in adversarial conditions become stack-level vulnerabilities. Microsoft’s unilateral cutoff of Nayara Energy’s cloud services is the clearest recent example of how a single foreign provider’s commercial or regulatory decision can impact an Indian strategic asset. Cursor under SpaceX creates the same exposure at the developer workflow layer, the layer where the code that runs critical systems actually gets written.
What India Should Do Differently
My own sense is that India is not powerless here, but the window to act from a position of negotiated strength is narrowing, and the Economic Survey 2025-26 said as much explicitly, warning that dependence on overseas AI platforms could expose India’s service-heavy economy to trade fragmentation and geopolitical disruptions. The leverage points are specific and actionable.
The first is building domestic AI coding tools rather than only consuming foreign ones. India’s IndiaAI Mission and startups like Sarvam AI show the beginnings of a domestic AI stack. The same institutional focus applied to developer tools, agentic IDEs, code models, and secure development infrastructure under Indian ownership, would create an alternative rather than a dependency. The second is applying the same security conditions to AI platforms that the government has already applied to Starlink. If foreign satellite communications infrastructure must route traffic through Indian gateways and submit to Indian regulatory oversight, AI platforms whose tools are used in critical systems should face analogous requirements: data localisation, model auditability, and crisis-mode controls that keep the kill switch in New Delhi rather than in California.
The third is segmenting critical workloads. Defence and strategic applications should run on sovereign stacks regardless of what general-purpose applications use. The most sensitive functions should not be architected on any single external provider, and Cursor under SpaceX is precisely the kind of provider whose single point of control should disqualify it from critical system development. The fourth is using public procurement as a policy instrument. Government and PSU purchasing can prioritise platforms that meet sovereignty and security criteria, creating demand for domestic tools and forcing foreign vendors to adapt to Indian norms rather than simply exporting global products at local prices.
The Honest Conclusion
There is a historical parallel here that India has lived through once before and cannot afford to repeat. India’s outsourcing boom of the 1990s and 2000s created extraordinary wealth and a world-class engineering talent base. It also locked India into the role of service provider for technology products designed, owned, and strategically directed elsewhere. India built the code. Others kept the IP, the margin, and the platform power.
The Cursor-SpaceX India push is the 2026 version of the same offer. More access to foreign tools at discount prices, in exchange for being the world’s largest generator of training data, usage behavioural signals, and market validation for a technology ecosystem whose strategic levers sit in Hawthorne, California. The short answer to whether India should use Cursor: individual developers will make their own choices, and the tool’s immediate productivity benefits are real.
The question India’s government and strategic community should be asking is not whether to use Cursor. It is whether India is building the domestic alternative that makes using Cursor a choice rather than a dependency. The distinction between those two positions is the difference between an AI power and an AI back office. India has not yet decided which it intends to be. Cursor’s India launch is a reminder that the decision cannot be deferred indefinitely.
Dr. Sudhanshu Kumar is a Subject Matter Expert on AI, Cyberwarfare and Cybersecurity at CENJOWS (Centre for Joint Warfare Studies), HQ (IDS), Ministry of Defence, New Delhi.
