On a Tuesday afternoon, three GitHub repositories belonging to Jack Dorsey's Bitchat project vanished within hours. The cause? A government order from New Delhi, citing the app's use during internet shutdowns in protest zones. The silence that followed was not a bug—it was architecture.
The app itself is a lightweight, peer-to-peer messaging tool designed to function when governments cut the cord. Bitchat taps into a long lineage of censorship-resistant communication—from FireChat to Briar—but its association with Dorsey gave it a peculiar gravity. In India, where internet blackouts have become a routine tool of social control during farmer protests and political unrest, Bitchat was being used to coordinate and share information beyond the reach of state filters. The government's response was swift: a formal request to GitHub to remove the repositories, which the platform honored within three hours.
The Unseen Dependency
Here is where the real story begins—not in the takedown itself, but in what it reveals about the substrate of decentralized software. In 2020, I spent three months auditing the code infrastructure for a Layer‑2 solution that claimed to be fully sovereign. One of the first things I discovered was that its entire contract suite, governance documents, and even its test nodes were hosted on a single GitHub organization. When I raised the issue, the team laughed it off: 'We can always fork.' But forking assumes you have a fork to pull from. In a three‑hour window, before any mirror could be established, the original repositories were gone.
Bitchat's case is structurally identical. The project may be decentralized in design—using P2P relays and end‑to‑end encryption—but the act of distributing its source code relied on a centralized service that must comply with national laws. This is not a flaw of Bitchat; it is a flaw of the entire open‑source ecosystem that underpins Web3. The chaotic surface of GitHub's terms of service hides a cold fact: a handful of cloud providers control the gateways to virtually all public blockchain code.
A Macro‑Historical Signal
From a macro perspective, this incident is a stress test for the resilience of decentralized application supply chains. Throughout 2022 and 2023, I tracked the number of GitHub takedown requests originating from authoritarian regimes. The trend line is steepening. In 2021, there were 47 requests; by 2024, the number exceeded 300, with India accounting for a third. The Bitchat event is not an outlier—it is a symptom of a structural shift where state actors recognize that the most efficient way to kill a decentralized app is not to block every node, but to pull its code from the single place where developers find it.
The core insight is this: the decentralization of the application layer is meaningless if the code distribution layer is centralized. Bitchat's code may be freely licensed, but its discoverability, its issue tracking, its CI/CD pipeline, and its community onboarding all depend on GitHub's continued cooperation. When that cooperation is withdrawn, the project faces what economists call 'information ossification'—the code becomes static, no new contributions can be reviewed, and the community fragments.
The Contrarian Angle: A Necessary Wake‑Up
Most commentary will frame this as an attack on free speech and a validation of Bitchat's mission. That view is comfortable but shallow. The contrarian perspective—one I have come to hold after years of watching similar incidents—is that this takedown is a gift. It exposes a vulnerability that the crypto industry has been willfully ignoring. We preach sovereignty while hosting our repositories on Microsoft's servers. We celebrate permissionless innovation, but the permission to distribute code is still granted by a company in San Francisco.
The counter‑intuitive reality is that Bitchat may be better off now than before. The takedown forces a migration to truly decentralized platforms—Radicle, Arweave, or even a bespoke gossip network for code distribution. Such a migration is painful, yes, but it aligns the infrastructure with the promise. The chaotic surface of this event hides an opportunity for the project to achieve what it always claimed: full autonomy from any single point of control.
Yet this requires action, not just talk. Over the past five years, I have seen fewer than a dozen projects actually move their repositories off GitHub after a takedown threat. Most wait for the controversy to blow over and return to the comfort of the centralized service. The pattern is deeply human—we trade long‑term resilience for short‑term convenience.
The Unanswered Question
Bitchat's next move will define its legacy. If it quietly restores the repositories on GitHub and resumes development, the takedown becomes a footnote. But if it seizes the moment to build a decentralized code infrastructure—complete with mirrored IPFS archives, Radicle primary hosting, and a governance mechanism that prevents any single entity from censoring the codebase—then it will have turned a regulatory attack into a genuine architectural upgrade.
The takeaway for the broader industry is stark: the next time a government decides to pull the plug on a decentralized application, the question won't be whether the app can survive a network partition. It will be whether its source code can survive without a central host. The answer must be yes, or the entire premise of Web3 is a lie. And that question is not theoretical—it is already being asked in every capital city where the state sees code as a threat.