By Sandeep Ajgaonkar
TBTF. Too Big to Fail.
The phrase took hold in the United States in 2008, when the subprime crisis, the collapse of Lehman Brothers and the rescues of Bear Stearns and AIG showed that some institutions were so interconnected their failure could hurt far more than their shareholders. Their failure was no longer a private matter.
Regulators responded with closer scrutiny, capital requirements and stress testing.
Is technology nearing a similar moment? Have some companies become Too Big to Fail or, more accurately, Too Embedded to Fail?
The real systemic risk in technology may no longer come from size. It comes from dependency.
Technology holds everything together:
Governments – cloud infrastructure.
Companies – collaboration, identity, cybersecurity and enterprise platforms.
Developers – code repositories and AI infrastructure.
Citizens – digital payments, communications and online public services.
The modern economy does not simply use technology. It runs on it.
When a bank fails, contagion is financial. When a critical platform fails, it is operational: a solvent organisation can find employees unable to communicate and customers unable to access services.
So, the question is no longer: How large is the company? It is: What stops working if the company stops working?
TBTF 2.0
Traditional TBTF analysis weighs size, interconnectedness, substitutability and complexity. Technology needs another lens, systemic digital dependency, built on five questions:
1. Dependency: How many critical economic or government functions rely on the platform?
2. Concentration: How easily can customers move elsewhere?
3. Interconnectedness: How many other services rely on its infrastructure?
4. Substitutability: How quickly could equivalent capability be restored?
5. Sovereignty: Would its failure, restriction or control affect another country’s ability to run critical national functions?
That last question is where technology departs from banking.
The sovereignty problem
The original TBTF debate was largely domestic. Technology does not fit inside those borders: many critical platforms are headquartered in the United States, but their infrastructure, customers and impact span the world.
If such a company runs critical infrastructure in India, Europe or Japan, but decisions are made elsewhere, whose systemic interest comes first during a crisis?
This is not an argument against Big Tech or for protectionism. It is recognition that the headquarters may be national, but the dependency is global.
What should India do?
India is one of the world’s largest digital societies. The answer cannot be localisation or isolation.
MeitY’s March 2026 cloud-selection framework already asks ministries to classify applications by sensitivity and criticality, with Category A covering systems such as Aadhaar, UPI, PAN, Passport and e-Courts.
The next step is to move from data sovereignty to dependency sovereignty. One asks: Where is our information?
The other: Can the system keep running if an external provider becomes unavailable? Five principles follow:
1. Identify systemically important technology providers: The test should be dependency, not revenue or market capitalisation.
2. Stress-test digital dependencies: Banks are stress-tested against extreme scenarios. Governments should know what happens if a major cloud, identity or AI provider fails. Not theoretically. Operationally.
3. Build genuine portability: Multi-cloud sounds reassuring on PowerPoint. The real test is whether critical workloads can move when required.
4. Create sovereign fallback capability: Not an Indian copy of every global platform, but credible alternatives that keep critical functions running.
5. Treat concentration as a board-level risk: When payments, communications and AI depend on a handful of providers, an outage becomes a customer crisis fast.
Too Big to Fail may be the wrong phrase
Some technology companies may not be too big to fail. They may be Too Connected, Too Embedded or, most importantly, Too Critical to Fail.
The distinction matters. Regulating size risks punishing success. Regulating systemic dependency asks: What happens to society when products or services suddenly disappear?
The biggest question is not whether Big Tech is Too Big to Fail, but this: If a company in one country becomes critical infrastructure for the world, who ensures it cannot fail?
And for India: How much of our digital future are we comfortable allowing someone else to decide?
-Sandeep Ajgaonkar is a former Editor of Express Computer. The views expressed are his own.