In October 2025, a single software update triggered the biggest shift in Bitcoin's node landscape in years. Bitcoin Core version 30 raised a limit on how much arbitrary data a transaction could carry, and within weeks more than a fifth of the network's node operators had switched to rival software rather than accept the change. Nobody could stop them. Nobody could overrule them. That is Bitcoin governance in action.
Bitcoin governance is the process by which the Bitcoin network decides on, adopts, or rejects changes to its rules, without any central authority in charge. There is no CEO, no board, no foundation with veto power. Instead, developers, miners, node operators, and users hold each other in check through a system often described as rough consensus.
This guide explains how that system works, who actually controls Bitcoin, how a Bitcoin Improvement Proposal becomes part of the protocol, and what the network's biggest governance battles reveal about where the real power sits.
Use the multichain Bitcoin.com Wallet app, trusted by millions to safely and easily send, receive, buy, sell, trade, and manage the most popular cryptocurrencies.
Key Takeaways
- Bitcoin has no formal governance body. Changes happen only when developers, miners, node operators, and economic users broadly agree, a standard known as rough consensus.
- No single group is in charge. Developers can write code but cannot force anyone to run it. Miners can build blocks but cannot change the rules. Nodes enforce the rules but cannot compel miners to follow a new ruleset.
- Formal changes are proposed through Bitcoin Improvement Proposals (BIPs), a public process anyone can participate in.
- Forks are Bitcoin's final vote. If the community cannot agree, the chain can split, as it did in 2017 when Bitcoin Cash emerged from the block size dispute.
- Governance was stress-tested in August 2026, when the contested BIP-110 soft fork attempted activation with 2.53% miner support. Its enforcing nodes split onto a minority chain that produced two blocks in eight hours and then stalled. The proposal is now marked Closed.
- The 21 million coin cap could be changed in theory, but only if nearly everyone agreed to devalue their own holdings, which is why it has never come close to happening.
What Is Bitcoin Governance?
Bitcoin governance is the set of processes through which participants in the Bitcoin network propose, debate, adopt, or reject changes to the protocol's rules. It covers everything from minor code improvements to consensus-level changes that redefine what counts as a valid block.
It helps to separate two ideas that often get blurred together:
- Governance by the protocol - Bitcoin's code already governs day-to-day operation automatically. It enforces the 21 million coin supply cap, adjusts mining difficulty every 2,016 blocks, and rejects invalid transactions without any human intervention.
- Governance of the protocol - This is the human layer: how people decide whether the code itself should change. Bitcoin handles this entirely off-chain, through public discussion, code review, and voluntary adoption, rather than through token-holder voting recorded on the blockchain.
This makes Bitcoin's governance model very different from a company, a central bank, or even most other blockchains. A central bank can change monetary policy with a committee vote. A software company can push an update to every user overnight. Bitcoin can do neither. Every participant chooses which software to run, and the network only functions because thousands of independent operators voluntarily run compatible rules.
The term most often used for how agreement forms is rough consensus, an idea borrowed from the engineering culture that built the early internet. Rough consensus does not mean unanimous approval or a formal vote. It means a proposal has been publicly examined, serious objections have been addressed or have failed to gain traction, and no significant constituency is prepared to block it.
Who Controls Bitcoin? The Five Power Centers
The short answer to "who controls Bitcoin" is: no one, and everyone. The longer answer is that several distinct groups each hold a specific, limited kind of power, and each can check the others. Satoshi Nakamoto designed the system this way, then reinforced the point by disappearing in 2011 and leaving no successor.
| Stakeholder | What they control | What they cannot do |
|---|---|---|
| Developers | Write, review, and maintain node software such as Bitcoin Core; propose improvements | Cannot force anyone to install or run their code |
| Miners | Assemble transactions into blocks; signal readiness for upgrades; direct hashrate | Cannot change consensus rules; invalid blocks are rejected by nodes no matter how much hashpower produced them |
| Node operators | Independently validate every block and transaction; choose which ruleset to enforce | Cannot produce blocks or compel miners to adopt new rules |
| Users and holders | Assign economic value; choose wallets, software, and which chain to treat as "Bitcoin" | Cannot directly alter code or block production |
| Businesses and exchanges | Decide which chain gets listed, traded, and integrated; concentrate large economic weight | Cannot override the rules users and nodes enforce |
A few points in this table deserve emphasis, because they are the most commonly misunderstood parts of Bitcoin's governance model.
Developers are proposers, not rulers - Bitcoin Core is the dominant node software, but it is one implementation of the protocol, not the protocol itself. Its maintainers merge code into a repository; they cannot install that code on a single machine they do not own. When enough operators dislike a release, they simply do not upgrade, or they switch to alternative software. That is exactly what happened after the version 30 release in 2025, covered below.
Miners follow rules, they do not set them - A common belief is that miners vote on Bitcoin's rules with their hashpower. In reality, miner signaling is a coordination tool for timing upgrades, not a ballot. If miners produced blocks that violated the consensus rules, every validating node on the network would reject those blocks automatically, and the miners would forfeit their block rewards. Even a majority of hashrate cannot change the protocol; a 51% attacker could disrupt transactions but could not print extra coins or alter the supply schedule.
Full nodes are the enforcement layer - Every full node independently checks every block against its own copy of the rules. As of early September 2026, public trackers counted roughly 23,800 reachable Bitcoin nodes worldwide, and the true number including private nodes is higher. This distribution is what makes unilateral rule changes practically impossible: there is no switch to flip, only tens of thousands of individual decisions.
Enforcement power has limits, though - Nodes can refuse to accept blocks, but refusing is not the same as governing. The August 2026 BIP-110 episode, covered below, showed what happens when a minority of nodes enforce a rule that miners will not build for: they get a technically valid chain that almost nobody mines, and it grinds to a halt. Governance requires the constituencies to align, not simply for one of them to dig in.
Economic weight decides contested outcomes - When governance disputes escalate, the deciding factor has historically been the economic majority: the users, businesses, and exchanges whose demand gives each chain its value. Hashpower follows price, because miners are paid in the coin they mine. A chain with rules that users reject quickly becomes unprofitable to mine.
How Is Bitcoin Governed in Practice? The BIP Process
Most protocol changes travel through a formal pipeline called the Bitcoin Improvement Proposal process. A Bitcoin Improvement Proposal (BIP) is a public design document that specifies a proposed change, the reasoning behind it, and its technical details. The system was introduced in 2011 by developer Amir Taaki, who modeled it on the proposal process used by the Python programming language, and it has structured Bitcoin development ever since. Every proposal, from the historic to the rejected, is archived in the public BIPs repository on GitHub.
BIPs fall into three categories:
- Standards Track BIPs propose changes that affect the network itself, such as new transaction formats or consensus rule changes.
- Informational BIPs document design guidelines or general information without proposing a rule change.
- Process BIPs propose changes to how Bitcoin development itself works, including the BIP process and upgrade activation methods.
The journey from idea to activated rule typically follows five stages:
- Discussion: Ideas surface informally on the public Bitcoin development mailing list, on GitHub, and at developer meetings. Many ideas die here, which is the system working as intended.
- Drafting: The author writes a formal specification: motivation, technical design, backward compatibility, and activation method. The draft receives a BIP number and enters public review.
- Peer review and consensus building: Developers and the wider community pick the proposal apart. Consensus-level changes usually also require a working reference implementation. This stage can take years; the Taproot upgrade spent almost four years between first proposal and activation.
- Implementation and release: If rough consensus emerges, the code is merged into node software such as Bitcoin Core and shipped in a release. Shipping code still changes nothing on its own.
- Activation: For consensus changes, the network coordinates a switch-on. The most common method is miner signaling: miners set a bit in the blocks they mine to indicate readiness, and once a threshold is met within a difficulty period, the new rules lock in. Taproot used a variant called Speedy Trial and locked in after more than 90% of blocks signaled support in mid-2021. Node operators and users then hold the final say by choosing whether to run the software that enforces the change.
The pipeline is deliberately slow and conservative. Bitcoin secures a network valued in the trillions of dollars at its 2025 peak, and a bugged consensus change could be catastrophic and irreversible. In Bitcoin governance, the default answer to any proposal is no.
Forks: Bitcoin's Final Vote
When the community genuinely cannot agree, the disagreement gets settled by forking, which is what happens when different participants run incompatible rulesets.
| Soft fork | Hard fork | |
|---|---|---|
| What changes | Rules get stricter; previously valid things may become invalid | Rules get looser or fundamentally different; previously invalid things become valid |
| Backward compatible? | Yes, non-upgraded nodes still accept new blocks | No, non-upgraded nodes reject new blocks |
| Network outcome | Network usually stays unified | Permanent chain split unless adoption is unanimous |
| Examples | SegWit (2017), Taproot (2021), attempted BIP-110 (2026) | Bitcoin Cash split (2017), BLAKE2b chain (2026) |
Soft forks tighten the rules in a backward-compatible way, which is why nearly every successful Bitcoin upgrade has used this route. Hard forks break compatibility and split the chain into two independent networks, each with its own coin.
A fork is the ultimate expression of Bitcoin's voluntary governance. Nobody can be forced onto a ruleset they reject, but nobody can force the market to value the chain they choose either. The 2017 Bitcoin Cash split proved both halves of that sentence at once, and the 2026 BIP-110 episode proved them again in compressed form: the dissenting minority successfully exercised its right to leave, and immediately discovered that leaving costs hashrate, liquidity, and exchange support.
Bitcoin Governance in Action: Three Defining Battles
Abstract descriptions only go so far. Bitcoin's governance model is best understood through the conflicts that tested it.
The Block Size War (2015 to 2017)
The most consequential dispute in Bitcoin's history was over a single parameter: the 1 MB block size limit. One camp wanted bigger blocks so Bitcoin could process more payments directly on-chain. The other camp argued that bigger blocks would make running a full node more expensive, eroding the decentralization that gives Bitcoin its value.
The fight escalated through failed compromise agreements between miners and businesses, competing software implementations, and years of public argument. Three outcomes settled it:
- SegWit activated in August 2017 as a soft fork, after a grassroots user-activated soft fork (UASF) movement threatened to have nodes reject blocks from miners who refused to signal for it. Miner signaling surged, and the upgrade locked in.
- Bitcoin Cash split off on August 1, 2017 as a hard fork with an 8 MB block size, giving the big-block camp its own chain rather than a change to Bitcoin's rules.
- SegWit2x collapsed in November 2017. A plan backed by a majority of hashrate and major companies to double the block size was cancelled days before activation, because node operators and users made clear they would not follow the new chain.
The lesson has shaped every governance debate since: neither miners nor corporations could impose a change on Bitcoin. The economic majority, expressed through the software users actually run, held the decisive power.
Taproot (2021): Consensus Done Smoothly
Taproot, Bitcoin's most recent major upgrade, showed the other side of the coin. Proposed in 2018, it improved privacy and enabled more efficient smart contracts, and it attracted almost no principled opposition. After years of review, miners signaled support above the 90% threshold in June 2021, and the upgrade activated at block 709,632 in November 2021. The contrast with the block size era demonstrated that Bitcoin's slow process can deliver significant change when rough consensus genuinely exists.
The Data Wars: Core v30, Knots, and BIP-110 (2023 to 2026)
Bitcoin's most recent governance battle was about what the blockchain is for, and it produced the most decisive rejection of a consensus change since 2017. The rise of Ordinals inscriptions and token schemes from 2023 onward filled blocks with images and arbitrary data, angering participants who see Bitcoin strictly as money and pleasing those who see permissionless blockspace as a feature.
The dispute produced a chain of governance events that will be studied for years:
| Date | Event | Governance significance |
|---|---|---|
| Oct 10, 2025 | Bitcoin Core v30 released, raising the default OP_RETURN data limit from 80 bytes to 100,000 bytes | Developers shipped a contested policy change after months of internal debate |
| Late 2025 | Node operators defected to Bitcoin Knots, an alternative implementation with strict data filters; Knots grew from roughly 2% to over 21% of reachable nodes by October 2025 | Users demonstrated the exit option: reject a release without leaving the network |
| Mar to Jul 2026 | BIP-110, a proposed one-year soft fork restricting arbitrary data, sought miner signals using a lowered 55% activation threshold instead of the traditional 95% | A minority attempted to change consensus rules without broad agreement |
| Jul 11, 2026 | Michael Saylor and Adam Back publicly opposed the proposal, arguing that invalidating fee-paying transactions set a worse precedent than the spam itself | The economic constituency weighed in against converting a policy dispute into a consensus rule |
| Aug 8, 2026 (block 961,632) | Mandatory signaling began. Only 51 of the prior 2,016 blocks had signaled support, or 2.53%. Enforcing nodes began rejecting non-signaling blocks and split onto a minority chain | The decisive test arrived, and the proposal's own threshold was missed by more than fifty percentage points |
| Aug 8 to 9, 2026 | The minority chain produced two blocks in roughly eight hours, then stalled, reaching only about four blocks in its first week while the main chain advanced normally | Roughly 99.85% of hashrate stayed on the original chain. Saylor estimated the branch would need around 25 years to reach its first difficulty adjustment |
| Aug 2026 | The BIPs repository marked BIP-110 as Closed, citing the chain split and stalled mining | The formal process recorded the rejection |
| Aug 30 to Sep 1, 2026 | Backers escalated to a hard fork replacing SHA-256d with BLAKE2b, an algorithm existing Bitcoin mining hardware cannot run. Luke Dashjr left mining pool OCEAN and now calls the main chain "Spamcoin" | Unable to win the constituencies, the minority chose to exit and build a separate network with its own asset |
Three aspects make this saga a governance case study worth studying closely.
First, the Core v30 backlash showed that even the dominant developer team faces real accountability. Operators switched software in the tens of percent rather than accept a default they disliked, and because Core and Knots enforce identical consensus rules, they did so without splitting the chain. Exit is available at the policy layer without breaking the network.
Second, BIP-110 demonstrated the limit of that same exit option once it reaches the consensus layer. A rule enforced by a few percent of nodes and effectively none of the hashrate does not change Bitcoin; it produces a separate chain nobody mines. Compare Taproot's +90% miner support with BIP-110's 2.53%, and the invisible requirement for genuine multi-constituency alignment becomes visible.
Third, the speed of the resolution is itself the finding. The Block Size War took roughly two years to settle the question of whether miners or nodes controlled consensus. BIP-110 settled its equivalent question in about eight hours. In 2017 the answer was genuinely unknown. By 2026 it was not: hashpower cannot force a rule change, and neither can a determined minority of node operators. Both need the other, plus the economic majority that gives either chain value.
The aftermath is the honest coda. Rather than accept the outcome, BIP-110's backers moved to a proof-of-work hard fork, switching to BLAKE2b so that existing SHA-256 miners are excluded by design. That is a legitimate use of open-source software and the clearest possible illustration of Bitcoin's voluntary nature: anyone who rejects the consensus can leave and start a network on their own terms. It also illustrates the cost. As of September 1, 2026, no major exchange, mainstream wallet, or Lightning implementation had publicly committed to supporting the new chain, and reviewers questioned whether enough compatible hardware would appear to sustain regular block production. A chain split does not divide Bitcoin's value evenly. It creates a new network that has to earn its own.
How Bitcoin's Governance Model Compares to Other Blockchains
Bitcoin's approach is only one of several governance models in use across crypto networks, and the differences matter.
| Bitcoin | Ethereum | On-chain governance networks (e.g., Tezos, Polkadot) | |
|---|---|---|---|
| Model | Off-chain rough consensus | Off-chain, with coordinating foundation | On-chain token voting |
| Who proposes changes | Anyone, via BIPs | Anyone, via Ethereum Improvement Proposals (EIPs) | Anyone, via formal on-chain proposals |
| Who decides | Developers, miners, nodes, and users through voluntary adoption | Core developer calls and community, with the Ethereum Foundation funding and coordinating | Token holders, weighted by stake |
| Speed of change | Very slow, years per major upgrade | Moderate, regular scheduled hard forks | Fast, protocol can amend itself automatically |
| Key risk | Stagnation, coordination failure | Perceived influence of a central foundation | Plutocracy: the wealthiest holders dominate votes |
| Key strength | Extreme resistance to capture | Balance of coordination and openness | Clear, fast, legitimized decision paths |
On-chain governance systems encode voting directly into the protocol: holders stake or delegate tokens, and approved proposals can execute automatically. That buys speed and clarity at the cost of concentrating power in large holders. Bitcoin sits at the opposite extreme. Its refusal to formalize decision-making is a deliberate design choice: any formal mechanism would create a target to capture, whether by governments, corporations, or whales. The trade-off is that Bitcoin changes very slowly, and its disputes are messy, public, and sometimes years long. Supporters argue that for a monetary network, this conservatism is the entire point.
Can Bitcoin's 21 Million Cap Be Changed?
Technically, yes. The supply cap is a few lines of code, and code can be edited. Practically, it is the closest thing in software to an immovable object.
Changing the cap would require a hard fork adopted by an overwhelming economic majority: developers to write it, miners to mine it, node operators to run it, and users to value the resulting coin. Every one of those groups holds bitcoin whose scarcity underpins its value, so the proposal asks nearly everyone to vote for their own dilution. Anyone who disagreed could simply keep running the old rules, and the 21 million coin chain would continue to exist. History suggests the market would treat that original chain as Bitcoin, just as it did in 2017.
The cap debate resurfaces periodically around the security budget question: as the block subsidy halves roughly every four years (most recently in April 2024), transaction fees must eventually fund network security. That is a genuine long-term design discussion. But it faces the same governance physics as everything else in Bitcoin: no rough consensus, no change.
Common Misconceptions About Bitcoin Governance
- "Bitcoin Core developers control Bitcoin." - They maintain the most widely used software, which gives them influence, not authority. The v30 episode, where over a fifth of node operators switched implementations within weeks, showed how quickly influence evaporates when operators disagree.
- "Miners vote on the rules." - Miner signaling coordinates upgrade timing; it is not a referendum. Nodes reject rule-breaking blocks regardless of the hashpower behind them.
- "A 51% attack could change Bitcoin's rules." - A majority of hashrate could censor or reorder recent transactions, which is damaging but temporary. It could not alter the supply cap, steal coins from addresses, or change consensus rules, because every full node validates independently.
- "Whoever owns the most bitcoin has the most say." - Holdings confer economic influence, not governance rights. There is no token voting in Bitcoin. A holder of one million coins runs the same rules as a holder of one satoshi.
- "No governance means no change." - Bitcoin has upgraded many times: pay-to-script-hash in 2012, SegWit in 2017, Taproot in 2021, among others. Change is possible; it simply requires broad, demonstrated agreement.
Conclusion
Bitcoin governance is the off-chain, consensus-driven process by which a leaderless network decides its own rules, balancing power between developers who propose, miners who build, nodes that enforce, and users who give the whole thing value. The system looks chaotic from the outside, and its disputes can drag on for years. Yet as of September 2026 it has kept a monetary network worth more than a trillion dollars running for over seventeen years without a CEO, a headquarters, or a single successful hostile takeover. The BIP-110 episode demonstrated the model's core property once more: no faction, however committed, changes Bitcoin's rules without the network's genuine agreement, and those who cannot accept that outcome are free to leave and build elsewhere. If you want to inspect the machinery yourself, every proposal ever made is public, starting with the original Bitcoin whitepaper and continuing through today's open debates.
