BSV preserves Satoshi\u2019s original approach to emergency network protection, while BTC Core abandoned the mechanism because of failures in its management and security.<\/li>\n<\/ul>\nTwo days ago, on October 31, 2026,<\/strong> the crypto press ran its annual White Paper Day routine: candle charts, a Saylor quote, \u201cwe are all Satoshi.\u201d Yesterday was a quieter anniversary. On November 1, 2016<\/strong>, 10 years ago, bitcoin.org posted a notice titled \u201cAlert System Retirement,\u201d and the one instrument Satoshi Nakamoto built so an honest network could say \u201cstop\u201d began its eighteen-month execution.<\/p>\nToday is All Souls Day, when the Western Christian Church remembers its dead and listens for them. The grave I have in mind holds an instrument, not a man.<\/p>\n
The alert key is the cleanest window we have into what BTC\u2018s maintainers decided Bitcoin is for. Satoshi built it under fire, in August 2010, ten days after he patched the worst bug in Bitcoin\u2019s history, as his own answer to a question every money system eventually faces: what does an honest network do when something goes badly wrong?<\/p>\n
Core answered that question differently. BSV answered it the way Satoshi did, and took the heat for it.<\/p>\n
<\/span>19 bad days<\/strong><\/span><\/h3>\nI wrote the summer of 2010 at length in Part 3 of my Written History of Bitcoin series at kurtwuckertjr.com, so if you want to learn about the chaos of that year, give it a read.<\/p>\n
On July 28, 2010<\/strong>, an OP_1 OP_RETURN prefix turned out to make any script evaluate true, so anyone could spend anyone\u2019s coins. Satoshi patched it within hours. On August 15, Jeff Garzik, one of the handful of developers then working alongside Satoshi, opened a bitcointalk thread called \u201cStrange block 74638\u201d because that block contained a transaction creating 184,467,440,737.09551616 BTC, two to the sixty-fourth satoshis. The sum had overflowed. Satoshi shipped v0.3.10 in about five hours, and the patched chain overtook the poisoned one at block 74,691.<\/p>\nCVE-2010-5139.<\/p>\n
Same day, same man, commit 4bd188c, message: \u201cmisc changes.\u201d That commit switched off roughly fifteen script opcodes (OP_CAT, OP_SUBSTR, OP_LSHIFT, OP_MUL and friends) as a denial-of-service precaution. BIP 347, the 2024 proposal to bring OP_CAT back to BTC, cites that exact commit by hash. BSV turned the opcodes back on in rolling updates between 2020 and 2026.<\/p>\n
A 14-year detour, filed under \u201cmisc.\u201d<\/p>\n
Ten days later, on August 25, 2010<\/strong>, came the commitment that gave the network a voice. Title: \u201calert system.\u201d Three safe-mode commits followed within the hour. It shipped in v0.3.11, and on August 27<\/strong>, Satoshi posted \u201cVersion 0.3.11 with upgrade alerts\u201d to bitcointalk.<\/p>\nThree consensus-grade changes in 19 days, under fire, by one guy!<\/p>\n
The fourth was the one that could talk to all nodes with authority.<\/p>\n
<\/span>What he actually built<\/strong><\/span><\/h3>\nHis release post is tiny. \u201cThe alert system can display notifications on the status bar.\u201d A line of text in your client. \u201cIn response to an alert, your node may also go into safe mode,\u201d and safe mode disables the JSON-RPC send and balance commands \u201cto protect it from losing money until you get a chance to upgrade.\u201d If you disagreed with him, \u201cyou can use the switch -disablesafemode to re-enable them.\u201d And the sentence the kill-switch crowd never quotes: \u201cNodes keep operating and do not stop generating in response to an alert.\u201d<\/p>\n
So the dreaded centralized override froze your wallet\u2019s RPC send command, left mining untouched, and could be waved off with one flag. The purpose he gave was fork hygiene: the alert system \u201ccan make sure users are warned not to act on anything in the fork.\u201d Even that much was gone by v0.3.20, when safe mode on alert was removed.<\/p>\n
Garzik replied the same evening that alerts \u201care only accepted if signed by a specific key (held by satoshi)\u201d and \u201ccannot be generated by average nodes.\u201d Did the cypherpunks riot? Did anyone call it a backdoor?<\/p>\n
There was some bickering about it, but the answer is largely no, and one of the longer fights in that thread was about a compiler flag!<\/p>\n
Two weeks earlier, on August 11, 2010<\/strong>, Satoshi was in a thread called \u201cEscrow\u201d defending the dullest possible escrow design, where \u201ccustomer loses, but the thief doesn\u2019t get the money either.\u201d He wrote: \u201cImagine if gold turned to lead when stolen. If the thief gives it back, it turns to gold again.\u201d He added that \u201cmoney burning\u201d was the wrong phrase because the funds were never truly destroyed.<\/p>\nFourteen days after typing that, he built the signed network-wide warning, and the gold line comes back when this story reaches the BSV freeze. Stolen value going inert and coming back, and an honest network telling itself the truth in an emergency: those are not two topics.<\/p>\n
Back to the top \u2191<\/em><\/p>\n<\/span>3 for 3<\/strong><\/span><\/h3>\nSatoshi never sent an alert. He left the project before he had to.<\/p>\n
The first real alert went out on February 18, 2012<\/strong>, ahead of the February 20<\/strong> protocol change. On March 12, 2013<\/strong>, during the 0.8 database fork, alert 1032 went out at priority 5000: \u201cURGENT: chain fork, stop mining on version 0.8.\u201d Two chains, a split network, healed inside a day, and the alert key is a large part of why. The last real alert was April 11, 2014<\/strong>, the Heartbleed upgrade notice.<\/p>\nThree uses, three successes, and not one false alarm or rogue message, ever! The known holders were Satoshi, Gavin Andresen, and theymos, the forum administrator.<\/p>\n
Remember its successes, because the key was not killed for anything it did.<\/p>\n
Back to the top \u2191<\/em><\/p>\n<\/span>Where it went<\/strong><\/span><\/h3>\nOn September 10, 2016<\/strong>, on the bitcoin-dev list, Gregory Maxwell, Blockstream co-founder and Core contributor, wrote: \u201cIt was provided to MTGox by a developer.\u201d (A developer. Singular. Unnamed then, unnamed since.) Then: \u201cMTGox\u2019s systems\u2019 were compromised and later their CEO\u2019s equipment taken by the Japanese police.\u201d Mt. Gox\u2018s CEO was Mark Karpel\u00e8s, arrested on August 1, 2015<\/strong>. So the key that could put a message on every Bitcoin node on earth sat on the equipment of the CEO of a compromised exchange, and then in a Tokyo evidence room.<\/p>\nWhich developer handed it over? The thread does not say, and a decade later nobody has said\u2026<\/p>\n
Maxwell himself admitted, \u201cNo one can tell which of the key holders created a message.\u201d Wladimir van der Laan, Core\u2019s lead maintainer at the time: \u201cno one can know who it has spread to over all those years.\u201d Johnson Lau, another Core contributor: \u201cWe don\u2019t know the actual date because the key is shared by many people.\u201d<\/p>\n
The maintainers of the most valuable open-source project in history could not tell you how many copies of their emergency key existed, or who had them.<\/p>\n
Not exactly a clean chain of custody.<\/p>\n
Back to the top \u2191<\/em><\/p>\n<\/span>The killing<\/strong><\/span><\/h3>\nThen came November 1, 2016<\/strong> and the retirement notice, and bitcoin.org gave its reasons in its own words. The alert system \u201cis a source of centralization\u201d because \u201ca single Alert Key can send network-wide alerts.\u201d<\/p>\n\u201cAs more developers join, the Alert Key is given to others, but cannot be taken away from those who have left.\u201d \u201cThis has led to the Alert Key potentially falling into the hands of malicious actors.\u201d And it was no longer useful anyway, because users can get their news from major Bitcoin outlets (the press, in other words).<\/p>\n
Every one of those sentences describes a custody failure by the people writing them. The key did not centralize itself. Nobody forced a developer to copy it to an exchange. And the proposed replacement for a cryptographically signed message from the protocol\u2019s steward was checking Twitter\u2026<\/p>\n
From there, it runs like a demolition schedule. 2 November 2016: a pre-final alert. January 19, 2017:<\/strong> the final alert, which disabled the system and left every old node displaying a static message, \u201cAlert Key Compromised.\u201d March 8, 2017:<\/strong> Bitcoin Core 0.14 ships with that final alert hard-coded for legacy nodes. May 1, 2017:<\/strong> the planned release of the key is postponed, because \u201cOlder clients may contain Alert handling code which is exploitable using the alert key.\u201d The code itself was already gone from Core 0.13.0.<\/p>\nThen on July 3, 2018<\/strong>, Bryan Bishop and Andrew Chow, two Core developers, published the private key, mainnet and testnet, on bitcoin.org. The stated reason was the key\u2019s \u201cunknown dissemination and proliferation.\u201d<\/p>\nThe logic is airtight in its way: if anyone can sign, no signature means anything.<\/p>\n
The alert key was the network\u2019s one channel for an authoritative warning, and the only instrument through which an author long gone from the project could ever have addressed every node at once. Core did not retire that channel. It published the key, so that any future message signed with it is indistinguishable from an impostor\u2019s. That is not decommissioning a feature.<\/p>\n
That is salting a grave.<\/p>\n
Back to the top \u2191<\/em><\/p>\nPeople I respect think a signed message makes BSV a permissioned network, and yes, it is controversial.<\/p>\n
SV Node 1.1.0 reintroduced an alert system in early 2024, announced by\u00a0BSV Association alongside the Network Access Rules (NAR) and Digital Asset Recovery (DAR). The Association\u2019s hub says it takes \u201cits name and general concept from the original Bitcoin alert key\u201d but is \u201carchitecturally distinct\u201d: a separate libp2p publish-subscribe topic, not the P2P relay. Informational messages are software updates and technical notices, which is exactly what Satoshi shipped in 0.3.11.<\/p>\n
Core\u2019s own reasons, one at a time. One key? BSV has five holders and a threshold. Cannot be taken away from those who left? Named parties under a published rule set, not a file that wandered onto an exchange server. The BSV design is not a copy of the 2010 key. It is a point-by-point answer to the way the 2010 key failed, which is more than the people who buried it ever attempted.<\/p>\n
And what about freezing coins using a directive from NAR\/DAR? Remember the escrow thread of August 11, 2010<\/strong>: gold turns to lead when stolen, and if the thief gives it back, it turns to gold again. A frozen output is lead, and a reassigned output under a court order is the gold coming back. Satoshi did not write that line about an alert system. He wrote it two weeks before he built one, defending a design where the thief ends up with nothing, but the victim can be made whole again.\u00a0<\/p>\nYou can dislike where BSV got the idea, but you cannot call the idea foreign to Bitcoin or its creator: Satoshi Nakamoto.<\/p>\n
Back to the top \u2191<\/em><\/p>\n<\/span>The verdict<\/strong><\/span><\/h3>\nStrip the branding off.<\/p>\n
Satoshi built a signed, network-wide warning himself, under fire, in the worst month of Bitcoin\u2019s history. The key worked every time it was used, and it was killed for what it might do, after the people killing it had let it leak. And one chain decided the answer to \u201cwhat happens when the money is stolen\u201d was to publish the key and tell you to read the news, while the other rebuilt the instrument with the failure modes patched and a judge in the loop.<\/p>\n
Core\u2019s Bitcoin is the one where the author\u2019s only emergency instrument lies buried under salted ground. BSV\u2019s is the one where his actual answer to the emergency question is still running code, five signatures, and a court order away. Controversial? Sure. But it is All Souls Day, and I know which chain is still listening for a dead instrument and the idea it carried, and which one salted the ground so it would never have to.<\/p>\n
Back to the top \u2191<\/em><\/p>\nFAQs:<\/strong><\/p>\nWhat was Bitcoin\u2019s original alert system?<\/strong>Why did Satoshi create the alert system?<\/strong>Why did BTC Core retire the alert system?<\/strong>What happened to Bitcoin\u2019s original alert key?<\/strong>How does BSV\u2019s alert system differ from the original?<\/strong>This opinion piece is published to encourage discussion. The author\u2019s views are their own and do not constitute legal, procurement, or policy advice, nor do they represent the positions of CoinGeek or its partners.<\/em><\/p>\n\n<\/span>Related articles<\/span><\/h2>\n<\/section>\nBack to the top \u2191<\/em><\/p>\nWatch | DigiCon 2026: Interoperability starts with people<\/p>\n\n